├── .dev_scripts └── gather_models.py ├── .gitattributes ├── .github ├── CONTRIBUTING.md ├── ISSUE_TEMPLATE │ ├── 1-bug-report.yml │ ├── 2-feature-request.yml │ ├── 3-new-model.yml │ ├── 4-documentation.yml │ └── config.yml ├── pull_request_template.md └── workflows │ ├── lint.yml │ ├── merge_stage_test.yml │ ├── pr_stage_test.yml │ ├── publish-to-pypi.yml │ ├── scripts │ └── get_mmcv_var.sh │ └── test_mim.yml ├── .gitignore ├── .pre-commit-config-zh-cn.yaml ├── .pre-commit-config.yaml ├── .pylintrc ├── .readthedocs.yml ├── CITATION.cff ├── LICENSE ├── MANIFEST.in ├── README.md ├── README_mmrotate.md ├── README_zh-CN.md ├── configs ├── _base_ │ ├── datasets │ │ ├── dior.py │ │ ├── dota.py │ │ ├── dota_coco.py │ │ ├── dota_ms.py │ │ ├── dota_qbox.py │ │ ├── dotav15.py │ │ ├── dotav2.py │ │ ├── hrsc.py │ │ ├── hrsid.py │ │ ├── rsdd.py │ │ ├── semi_visdronezsd.py │ │ ├── srsdd.py │ │ ├── ssdd.py │ │ └── visdronezsd.py │ ├── default_runtime.py │ ├── models │ │ └── oriented-rcnn-le90_r50_fpn.py │ └── schedules │ │ ├── schedule_1x.py │ │ ├── schedule_3x.py │ │ ├── schedule_40e.py │ │ └── schedule_6x.py ├── cfa │ ├── README.md │ ├── cfa-qbox_r50_fpn_1x_dota.py │ ├── cfa-qbox_r50_fpn_40e_dota.py │ └── metafile.yml ├── convnext │ ├── README.md │ ├── metafile.yml │ └── rotated-retinanet-rbox-le90_convnext-tiny_fpn_kld-stable_adamw-1x_dota.py ├── csl │ ├── README.md │ ├── metafile.yml │ └── rotated-retinanet-rbox-le90_r50_fpn_csl-gaussian_amp-1x_dota.py ├── gliding_vertex │ ├── README.md │ ├── gliding-vertex-qbox_r50_fpn_1x_dota.py │ ├── gliding-vertex-rbox_r50_fpn_1x_dota.py │ └── metafile.yml ├── gwd │ ├── README.md │ ├── metafile.yml │ ├── rotated-retinanet-hbox-oc_r50_fpn_gwd_1x_dota.py │ ├── rotated-retinanet-rbox-le135_r50_fpn_gwd_1x_dota.py │ └── rotated-retinanet-rbox-le90_r50_fpn_gwd_1x_dota.py ├── h2rbox │ ├── README.md │ ├── dior │ │ └── h2rbox-le90_r50_fpn_adamw-1x_dior.py │ ├── dotav15 │ │ ├── h2rbox-le90_r50_fpn_adamw-1x_dotav15.py │ │ └── h2rbox-le90_r50_fpn_adamw-3x_dotav15.py │ ├── dotav2 │ │ ├── h2rbox-le90_r50_fpn_adamw-1x_dotav2.py │ │ └── h2rbox-le90_r50_fpn_adamw-3x_dotav2.py │ ├── h2rbox-le90_r50_fpn_adamw-1x_dota-ms.py │ ├── h2rbox-le90_r50_fpn_adamw-1x_dota.py │ ├── h2rbox-le90_r50_fpn_adamw-3x_dota.py │ └── metafile.yml ├── h2rbox_v2 │ ├── README.md │ ├── h2rbox_v2-le90_r50_fpn-1x_dota.py │ ├── h2rbox_v2-le90_r50_fpn-1x_dotav15.py │ ├── h2rbox_v2-le90_r50_fpn-1x_dotav2.py │ ├── h2rbox_v2-le90_r50_fpn-6x_hrsc.py │ ├── h2rbox_v2-le90_r50_fpn_ms_rr-1x_dota.py │ ├── h2rbox_v2-le90_r50_fpn_rr-6x_hrsc.py │ └── metafile.yml ├── kfiou │ ├── README.md │ ├── metafile.yml │ ├── r3det-oc_r50_fpn_kfiou-ln_1x_dota.py │ ├── roi-trans-le90_r50_fpn_kfiou-ln_1x_dota.py │ ├── rotated-retinanet-hbox-le135_r50_fpn_kfiou_1x_dota.py │ ├── rotated-retinanet-hbox-le90_r50_fpn_kfiou_1x_dota.py │ ├── rotated-retinanet-hbox-oc_r50_fpn_kfiou_1x_dota.py │ └── s2anet-le135_r50_fpn_kfiou-ln_1x_dota.py ├── kld │ ├── README.md │ ├── metafile.yml │ ├── r3det-oc_r50_fpn_kld-stable_1x_dota.py │ ├── r3det-oc_r50_fpn_kld_1x_dota.py │ ├── r3det-tiny-oc_r50_fpn_kld_1x_dota.py │ ├── rotated-retinanet-hbox-oc_r50_fpn_kld-stable_1x_dota.py │ ├── rotated-retinanet-hbox-oc_r50_fpn_kld-stable_rr-6x_hrsc.py │ ├── rotated-retinanet-hbox-oc_r50_fpn_kld_1x_dota.py │ ├── rotated-retinanet-rbox-le90_r50_fpn_kld-stable_1x_dota.py │ ├── rotated-retinanet-rbox-le90_r50_fpn_kld-stable_adamw-1x_dota.py │ ├── rotated-retinanet-rbox-le90_r50_fpn_kld-stable_rr-6x_hrsc.py │ └── rotated-retinanet-rbox-le90_r50_fpn_kld_1x_dota.py ├── oriented_rcnn │ ├── README.md │ ├── metafile.yml │ ├── oriented-rcnn-le90_r50_fpn_1x_dota.py │ ├── oriented-rcnn-le90_r50_fpn_6x_hrsid.py │ ├── oriented-rcnn-le90_r50_fpn_6x_rsdd.py │ ├── oriented-rcnn-le90_r50_fpn_6x_srsdd.py │ ├── oriented-rcnn-le90_r50_fpn_6x_ssdd.py │ ├── oriented-rcnn-le90_r50_fpn_amp-1x_dota.py │ └── oriented-rcnn-le90_swin-tiny_fpn_1x_dota.py ├── oriented_reppoints │ ├── README.md │ ├── metafile.yml │ ├── oriented-reppoints-qbox_r50_fpn_1x_dota.py │ └── oriented-reppoints-qbox_r50_fpn_mstrain-40e_dota.py ├── psc │ ├── README.md │ ├── metafile.yml │ ├── rotated-fcos-hbox-le90_r50_fpn_psc-dual_1x_dota.py │ ├── rotated-fcos-hbox-le90_r50_fpn_psc_rr-6x_hrsc.py │ ├── rotated-retinanet-rbox-le90_r50_fpn_psc-dual_amp-1x_dota.py │ └── rotated-retinanet-rbox-le90_r50_fpn_psc_rr-6x_hrsc.py ├── r3det │ ├── README.md │ ├── metafile.yml │ ├── r3det-oc_r50_fpn_1x_dota.py │ ├── r3det-refine-oc_r50_fpn_1x_dota.py │ └── r3det-tiny-oc_r50_fpn_1x_dota.py ├── redet │ ├── README.md │ ├── metafile.yml │ ├── redet-le90_re50_refpn_1x_dota.py │ ├── redet-le90_re50_refpn_3x_hrsc.py │ ├── redet-le90_re50_refpn_amp-1x_dota.py │ └── redet-le90_re50_refpn_rr-1x_dota-ms.py ├── roi_trans │ ├── README.md │ ├── metafile.yml │ ├── roi-trans-le135_r50_fpn_1x_dota.py │ ├── roi-trans-le90_r50_fpn_1x_dota-ms.py │ ├── roi-trans-le90_r50_fpn_1x_dota.py │ ├── roi-trans-le90_r50_fpn_amp-1x_dota.py │ ├── roi-trans-le90_r50_fpn_rr-1x_dota-ms.py │ ├── roi-trans-le90_swin-tiny_fpn_1x_dota.py │ └── roi-trans-oc_r50_fpn_1x_dota.py ├── rotated_atss │ ├── README.md │ ├── metafile.yml │ └── rotated-atss-le90_r50_fpn_1x_dota.py ├── rotated_faster_rcnn │ ├── README.md │ ├── metafile.yml │ └── rotated-faster-rcnn-le90_r50_fpn_1x_dota.py ├── rotated_fcos │ ├── README.md │ ├── metafile.yml │ ├── rotated-fcos-hbox-le90_r50_fpn_1x_dota.py │ ├── rotated-fcos-hbox-le90_r50_fpn_csl-gaussian_1x_dota.py │ ├── rotated-fcos-le90_r50_fpn_1x_dota.py │ ├── rotated-fcos-le90_r50_fpn_kld_1x_dota.py │ └── rotated-fcos-le90_r50_fpn_rr-6x_hrsc.py ├── rotated_reppoints │ ├── README.md │ ├── metafile.yml │ └── rotated-reppoints-qbox_r50_fpn_1x_dota.py ├── rotated_retinanet │ ├── README.md │ ├── metafile.yml │ ├── rotated-retinanet-hbox-le135_r50_fpn_1x_dota.py │ ├── rotated-retinanet-hbox-le90_r50_fpn_1x_dota.py │ ├── rotated-retinanet-hbox-oc_r50_fpn_1x_dota.py │ ├── rotated-retinanet-hbox-oc_r50_fpn_rr-6x_hrsc.py │ ├── rotated-retinanet-rbox-le135_r50_fpn_1x_dota.py │ ├── rotated-retinanet-rbox-le90_r50_fpn_1x_dior.py │ ├── rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py │ ├── rotated-retinanet-rbox-le90_r50_fpn_amp-1x_dota.py │ ├── rotated-retinanet-rbox-le90_r50_fpn_rr-1x_dota-ms.py │ ├── rotated-retinanet-rbox-le90_r50_fpn_rr-6x_hrsc.py │ ├── rotated-retinanet-rbox-oc_r50_fpn_1x_dior.py │ └── rotated-retinanet-rbox-oc_r50_fpn_1x_dota.py ├── rotated_rtmdet │ ├── README.md │ ├── _base_ │ │ ├── default_runtime.py │ │ ├── dota_rr.py │ │ ├── dota_rr_ms.py │ │ ├── hrsc_rr.py │ │ └── schedule_3x.py │ ├── metafile.yml │ ├── rotated_rtmdet_l-100e-aug-dota.py │ ├── rotated_rtmdet_l-300e-aug-hrsc.py │ ├── rotated_rtmdet_l-3x-dota.py │ ├── rotated_rtmdet_l-3x-dota_ms.py │ ├── rotated_rtmdet_l-9x-hrsc.py │ ├── rotated_rtmdet_l-coco_pretrain-3x-dota_ms.py │ ├── rotated_rtmdet_m-3x-dota.py │ ├── rotated_rtmdet_m-3x-dota_ms.py │ ├── rotated_rtmdet_s-3x-dota.py │ ├── rotated_rtmdet_s-3x-dota_ms.py │ ├── rotated_rtmdet_tiny-300e-aug-hrsc.py │ ├── rotated_rtmdet_tiny-3x-dota.py │ ├── rotated_rtmdet_tiny-3x-dota_ms.py │ └── rotated_rtmdet_tiny-9x-hrsc.py ├── s2anet │ ├── README.md │ ├── cascade-s2anet-le135_r50_fpn_1x_dota.py │ ├── metafile.yml │ ├── s2anet-le135_r50_fpn_1x_dota.py │ ├── s2anet-le135_r50_fpn_3x_hrsc.py │ ├── s2anet-le135_r50_fpn_amp-1x_dota.py │ └── s2anet-le90_r50_fpn_1x_dota.py └── sasm_reppoints │ ├── README.md │ ├── metafile.yml │ └── sasm-reppoints-qbox_r50_fpn_1x_dota.py ├── demo ├── MMRotate_Tutorial.ipynb ├── demo.jpg ├── dota_demo.jpg ├── huge_image_demo.py └── image_demo.py ├── docker ├── Dockerfile └── serve │ ├── Dockerfile │ ├── config.properties │ └── entrypoint.sh ├── docs ├── en │ ├── _static │ │ ├── css │ │ │ └── readthedocs.css │ │ └── image │ │ │ └── mmrotate-logo.png │ ├── advanced_guides │ │ ├── conventions.md │ │ ├── customize_dataset.md │ │ ├── customize_losses.md │ │ ├── customize_models.md │ │ ├── customize_runtime.md │ │ ├── customize_transforms.md │ │ ├── data_flow.md │ │ ├── datasets.md │ │ ├── engine.md │ │ ├── evaluation.md │ │ ├── how_to.md │ │ ├── index.rst │ │ ├── models.md │ │ └── transforms.md │ ├── api.rst │ ├── conf.py │ ├── get_started.md │ ├── index.rst │ ├── make.bat │ ├── migration.md │ ├── model_zoo.md │ ├── notes │ │ ├── changelog.md │ │ ├── changelog_v0.x.md │ │ ├── contribution_guide.md │ │ ├── faq.md │ │ └── projects.md │ ├── overview.md │ ├── stat.py │ ├── switch_language.md │ └── user_guides │ │ ├── config.md │ │ ├── dataset_prepare.md │ │ ├── deploy.md │ │ ├── index.rst │ │ ├── inference.md │ │ ├── label_studio.md │ │ ├── test_results_submission.md │ │ ├── train_test.md │ │ ├── useful_tools.md │ │ └── visualization.md └── zh_cn │ ├── _static │ ├── css │ │ └── readthedocs.css │ └── image │ │ └── mmrotate-logo.png │ ├── advanced_guides │ ├── conventions.md │ ├── customize_dataset.md │ ├── customize_losses.md │ ├── customize_models.md │ ├── customize_runtime.md │ ├── customize_transforms.md │ ├── data_flow.md │ ├── datasets.md │ ├── engine.md │ ├── evaluation.md │ ├── how_to.md │ ├── index.rst │ ├── models.md │ └── transforms.md │ ├── api.rst │ ├── conf.py │ ├── get_started.md │ ├── index.rst │ ├── make.bat │ ├── migration.md │ ├── model_zoo.md │ ├── notes │ ├── contribution_guide.md │ ├── faq.md │ └── projects.md │ ├── overview.md │ ├── stat.py │ ├── switch_language.md │ └── user_guides │ ├── config.md │ ├── dataset_prepare.md │ ├── deploy.md │ ├── index.rst │ ├── inference.md │ ├── label_studio.md │ ├── test_results_submission.md │ ├── train_test.md │ ├── useful_tools.md │ └── visualization.md ├── mmrotate ├── __init__.py ├── apis │ ├── __init__.py │ └── inference.py ├── datasets │ ├── __init__.py │ ├── dior.py │ ├── dota.py │ ├── hrsc.py │ ├── instance_dataset.py │ ├── nwpu45.py │ └── transforms │ │ ├── __init__.py │ │ ├── loading.py │ │ └── transforms.py ├── engine │ ├── __init__.py │ └── hooks │ │ ├── __init__.py │ │ └── mean_teacher_hook.py ├── evaluation │ ├── __init__.py │ ├── functional │ │ ├── __init__.py │ │ └── mean_ap.py │ └── metrics │ │ ├── __init__.py │ │ ├── dota_metric.py │ │ └── rotated_coco_metric.py ├── models │ ├── __init__.py │ ├── backbones │ │ ├── __init__.py │ │ └── re_resnet.py │ ├── dense_heads │ │ ├── __init__.py │ │ ├── angle_branch_retina_head.py │ │ ├── cfa_head.py │ │ ├── h2rbox_head.py │ │ ├── h2rbox_v2_head.py │ │ ├── oriented_reppoints_head.py │ │ ├── oriented_rpn_head.py │ │ ├── r3_head.py │ │ ├── rhino_align_head.py │ │ ├── rhino_head.py │ │ ├── rhino_ph_head.py │ │ ├── rhino_phc_head.py │ │ ├── rotated_atss_head.py │ │ ├── rotated_conditional_detr_head.py │ │ ├── rotated_dab_detr_head.py │ │ ├── rotated_deformable_detr_head.py │ │ ├── rotated_detr_head.py │ │ ├── rotated_fcos_head.py │ │ ├── rotated_reppoints_head.py │ │ ├── rotated_retina_head.py │ │ ├── rotated_rtmdet_head.py │ │ ├── s2a_head.py │ │ └── sam_reppoints_head.py │ ├── detectors │ │ ├── __init__.py │ │ ├── h2rbox.py │ │ ├── h2rbox_v2.py │ │ ├── refine_single_stage.py │ │ ├── rhino.py │ │ ├── rotated_dab_detr.py │ │ ├── rotated_deformable_detr.py │ │ ├── rotated_soft_teacher.py │ │ └── semi_base.py │ ├── layers │ │ ├── __init__.py │ │ ├── align.py │ │ └── transformer │ │ │ ├── __init__.py │ │ │ ├── rhino_layers.py │ │ │ ├── rhino_layers_v2.py │ │ │ ├── rotated_attention.py │ │ │ ├── rotated_dab_detr_layers.py │ │ │ ├── rotated_deformable_detr_layers.py │ │ │ └── utils.py │ ├── losses │ │ ├── __init__.py │ │ ├── convex_giou_loss.py │ │ ├── gaussian_dist_loss.py │ │ ├── gaussian_dist_loss_v1.py │ │ ├── h2rbox_consistency_loss.py │ │ ├── h2rbox_v2_consistency_loss.py │ │ ├── kf_iou_loss.py │ │ ├── rotated_iou_loss.py │ │ ├── smooth_focal_loss.py │ │ └── spatial_border_loss.py │ ├── necks │ │ ├── __init__.py │ │ └── re_fpn.py │ ├── roi_heads │ │ ├── __init__.py │ │ ├── bbox_heads │ │ │ ├── __init__.py │ │ │ ├── convfc_rbbox_head.py │ │ │ └── gv_bbox_head.py │ │ ├── gv_ratio_roi_head.py │ │ └── roi_extractors │ │ │ ├── __init__.py │ │ │ └── rotate_single_level_roi_extractor.py │ ├── task_modules │ │ ├── __init__.py │ │ ├── assigners │ │ │ ├── __init__.py │ │ │ ├── convex_assigner.py │ │ │ ├── dn_group_hungarian_assigner.py │ │ │ ├── match_cost.py │ │ │ ├── max_convex_iou_assigner.py │ │ │ ├── rotate_iou2d_calculator.py │ │ │ ├── rotated_atss_assigner.py │ │ │ └── sas_assigner.py │ │ ├── coders │ │ │ ├── __init__.py │ │ │ ├── angle_coder.py │ │ │ ├── delta_midpointoffset_rbbox_coder.py │ │ │ ├── delta_xywh_hbbox_coder.py │ │ │ ├── delta_xywh_qbbox_coder.py │ │ │ ├── delta_xywht_hbbox_coder.py │ │ │ ├── delta_xywht_rbbox_coder.py │ │ │ ├── distance_angle_point_coder.py │ │ │ └── gliding_vertex_coder.py │ │ └── prior_generators │ │ │ ├── __init__.py │ │ │ └── anchor_generator.py │ └── utils │ │ ├── __init__.py │ │ ├── enn.py │ │ ├── misc.py │ │ ├── orconv.py │ │ └── ripool.py ├── registry.py ├── structures │ ├── __init__.py │ └── bbox │ │ ├── __init__.py │ │ ├── bbox_overlaps.py │ │ ├── box_converters.py │ │ ├── quadri_boxes.py │ │ ├── rotated_boxes.py │ │ └── transforms.py ├── testing │ ├── __init__.py │ └── _utils.py ├── utils │ ├── __init__.py │ ├── collect_env.py │ ├── misc.py │ ├── patch │ │ ├── __init__.py │ │ ├── merge_results.py │ │ └── split.py │ └── setup_env.py ├── version.py └── visualization │ ├── __init__.py │ ├── local_visualizer.py │ └── palette.py ├── model-index.yml ├── projects ├── CastDetv2 │ ├── README.md │ ├── castdet │ │ ├── __init__.py │ │ ├── castdet.py │ │ ├── modified_resnet.py │ │ ├── ovd_bbox_head.py │ │ ├── pseudo_label_queue.py │ │ └── standard_roi_head2.py │ ├── configs │ │ ├── oriented-rcnn_r50-fpn_20k_visdronezsd_base-set.py │ │ └── visdrone_step2_castdet_12b_10k_oriented.py │ ├── resources │ │ ├── castdet.png │ │ ├── remoteCLIP_embeddings_bgs_normalized.npy │ │ ├── remoteCLIP_embeddings_normalized.npy │ │ ├── results_hbb.png │ │ ├── results_obb.png │ │ ├── toolbox.png │ │ ├── vis_result.png │ │ └── visdronezsd_split │ │ │ ├── test.txt │ │ │ ├── visdrone_initialize.txt │ │ │ ├── visdrone_labeled_3000.txt │ │ │ ├── visdrone_test.txt │ │ │ └── visdrone_unlabeled_8726.txt │ ├── run.sh │ └── tools │ │ ├── generate_text_embeddings.py │ │ └── merge_weights.py ├── GLIP │ ├── README.md │ ├── configs │ │ ├── glip_atss_r50_a_fpn_dyhead_visdronezsd_base.py │ │ ├── glip_atss_r50_a_fpn_dyhead_visdronezsd_base_nwpu.py │ │ └── glip_atss_r50_a_fpn_dyhead_visdronezsd_base_nwpu45_pseudo_labeling.py │ ├── glip │ │ ├── __init__.py │ │ └── atss_vlfusion_head.py │ └── run.sh ├── GroundingDINO │ ├── README.md │ ├── configs │ │ ├── grounding_dino_swin-t_visdrone_base-set_adamw.py │ │ ├── grounding_dino_swin-t_visdrone_base-set_adamw_nwpu45.py │ │ └── grounding_dino_swin-t_visdrone_base-set_adamw_nwpu45_pseudo_labeling.py │ ├── groundingdino │ │ ├── __init__.py │ │ ├── grounding_dino.py │ │ ├── grounding_dino_head.py │ │ └── grounding_dino_layers.py │ ├── run.sh │ └── tools │ │ └── merge_ovdg_preds.py ├── LSKNet │ ├── README.md │ ├── configs │ │ ├── lsk_s_ema_fpn_1x_dota_le90.py │ │ ├── lsk_s_fpn_1x_dota_le90.py │ │ ├── lsk_s_fpn_3x_hrsc_le90.py │ │ └── lsk_t_fpn_1x_dota_le90.py │ └── lsknet │ │ ├── __init__.py │ │ └── lsknet.py ├── LabelStudio │ └── backend_template │ │ ├── _mmrotate.py │ │ ├── _wsgi.py │ │ └── readme.md ├── RR360 │ ├── configs360 │ │ ├── .gitkeep │ │ ├── _base_ │ │ │ ├── datasets │ │ │ │ └── dota.py │ │ │ ├── default_runtime.py │ │ │ └── schedules │ │ │ │ ├── schedule_1x.py │ │ │ │ ├── schedule_3x.py │ │ │ │ └── schedule_6x.py │ │ ├── readme.md │ │ ├── rotated_retinanet │ │ │ └── rotated-retinanet-rbox-r360_r50_fpn_6x_ic19.py │ │ ├── rotated_rtmdet_x3 │ │ │ ├── _base_ │ │ │ │ ├── default_runtime.py │ │ │ │ ├── dota_rr.py │ │ │ │ └── schedule_3x.py │ │ │ └── rotated_rtmdet_l_l1-3x-ic19_pt.py │ │ └── rotated_rtmdet_x3_r │ │ │ ├── _base_ │ │ │ ├── default_runtime.py │ │ │ ├── dota_rr.py │ │ │ └── schedule_3x.py │ │ │ ├── rotated_rtmdet_l-3x-ic19_pt.py │ │ │ ├── rotated_rtmdet_l_l1-3x-ic19_pt.py │ │ │ ├── rotated_rtmdet_m_l1-3x-ic19_pt.py │ │ │ ├── rotated_rtmdet_s_l1-3x-ic19_pt.py │ │ │ └── rotated_rtmdet_tiny_l1-3x-ic19_pt.py │ ├── datasets │ │ ├── __init__.py │ │ └── transforms │ │ │ ├── __init__.py │ │ │ └── transforms.py │ ├── evaluation │ │ ├── __init__.py │ │ ├── functional │ │ │ ├── __init__.py │ │ │ └── mean_ap.py │ │ └── metrics │ │ │ ├── __init__.py │ │ │ └── dota_r360_metric.py │ ├── readme.md │ ├── readme_Zh.md │ ├── structures │ │ ├── __init__.py │ │ └── bbox │ │ │ ├── __init__.py │ │ │ └── rotated_boxes.py │ ├── tools │ │ ├── browse_dataset.py │ │ ├── test.py │ │ └── train.py │ └── visualization │ │ ├── __init__.py │ │ └── local_visualizer.py ├── ViLD │ ├── README.md │ ├── configs │ │ ├── vild_oriented-rcnn_r50_fpn_visdronezsd_step1_prepare.py │ │ ├── vild_oriented-rcnn_r50_fpn_visdronezsd_step2_finetune.py │ │ └── vild_visdronezsd.py │ ├── resources │ │ └── vild_framework.png │ ├── run.sh │ └── vild │ │ ├── __init__.py │ │ ├── modified_resnet.py │ │ ├── ovd_bbox_head.py │ │ └── rotated_vild.py └── example_project │ ├── README.md │ ├── configs │ └── r3det-oc_dummy-resnet_fpn_1x_dota.py │ └── dummy │ ├── __init__.py │ └── dummy_resnet.py ├── requirements.txt ├── requirements ├── build.txt ├── docs.txt ├── mminstall.txt ├── multimodal.txt ├── optional.txt ├── readthedocs.txt ├── runtime.txt └── tests.txt ├── resources ├── mmrotate-logo.png ├── qq_group_qrcode.jpg └── zhihu_qrcode.jpg ├── setup.cfg ├── setup.py ├── tests ├── data │ ├── dior │ │ ├── Annotations │ │ │ └── Oriented Bounding Boxes │ │ │ │ └── 00001.xml │ │ ├── JPEGImage │ │ │ └── 00001.jpg │ │ └── demo.txt │ ├── dota │ │ ├── images │ │ │ └── P2805__1024__0___0.png │ │ └── labelTxt │ │ │ └── P2805__1024__0___0.txt │ └── hrsc │ │ ├── FullDataSet │ │ └── Annotations │ │ │ └── 100000006.xml │ │ └── demo.txt ├── test_apis │ └── test_inference.py ├── test_datasets │ ├── test_dior.py │ ├── test_dota.py │ ├── test_hrsc.py │ └── test_transforms │ │ └── test_transforms.py ├── test_evaluation │ └── test_metrics │ │ ├── test_dota_metric.py │ │ └── test_rotated_coco_metric.py ├── test_models │ ├── test_backbones │ │ └── test_re_resnet.py │ ├── test_dense_heads │ │ ├── test_angle_branch_retina_head.py │ │ ├── test_cfa_head.py │ │ ├── test_h2rbox_head.py │ │ ├── test_oriented_reppoints.py │ │ ├── test_r3_head.py │ │ ├── test_rotated_atss_head.py │ │ ├── test_rotated_fcos_head.py │ │ ├── test_rotated_reppoints.py │ │ ├── test_rotated_retina_head.py │ │ ├── test_rotated_rtmdet_head.py │ │ ├── test_s2a_head.py │ │ └── test_sam_reppoints_head.py │ ├── test_detectors │ │ ├── test_h2rbox.py │ │ ├── test_refine_single_stage.py │ │ ├── test_single_stage.py │ │ └── test_two_stage.py │ ├── test_losses │ │ └── test_loss.py │ ├── test_necks │ │ └── test_re_fpn.py │ ├── test_roi_heads │ │ ├── test_bbox_head │ │ │ ├── test_convfc_rbbox_head.py │ │ │ └── test_gv_bbox_head.py │ │ └── test_gv_ratio_roi_head.py │ └── test_task_modules │ │ ├── test_assigners │ │ ├── test_convex_assigner.py │ │ ├── test_max_convex_iou_assigner.py │ │ ├── test_rotated_atss_assigner.py │ │ └── test_sas_assigner.py │ │ ├── test_coder │ │ ├── test_angle_coder.py │ │ ├── test_delta_midpointoffset_rbbox_coder.py │ │ ├── test_delta_xywh_hbbox_coder.py │ │ ├── test_delta_xywh_qbbox_coder.py │ │ ├── test_delta_xywht_hbbox_coder.py │ │ ├── test_delta_xywht_rbbox_coder.py │ │ └── test_gliding_vertex_coder.py │ │ ├── test_prior_generators │ │ └── test_anchor_generator.py │ │ └── test_rotated_iou2d_calculator.py ├── test_structures │ └── test_bbox │ │ ├── test_box_converters.py │ │ ├── test_quadri_boxes.py │ │ └── test_rotated_boxes.py └── test_visualization │ ├── test_local_visualizer.py │ └── test_palette.py └── tools ├── analysis_tools ├── analyze_logs.py ├── benchmark.py ├── browse_dataset.py ├── confusion_matrix.py └── get_flops.py ├── data ├── README.md ├── dior │ └── README.md ├── dota │ ├── README.md │ ├── dota2coco.py │ └── split │ │ ├── img_split.py │ │ └── split_configs │ │ ├── ms_test.json │ │ ├── ms_train.json │ │ ├── ms_trainval.json │ │ ├── ms_val.json │ │ ├── ss_test.json │ │ ├── ss_train.json │ │ ├── ss_trainval.json │ │ └── ss_val.json ├── hrsc │ └── README.md ├── hrsid │ └── README.md ├── rsdd │ └── README.md ├── srsdd │ └── README.md └── ssdd │ └── README.md ├── deployment ├── mmrotate2torchserve.py └── mmrotate_handler.py ├── dist_test.sh ├── dist_train.sh ├── misc └── print_config.py ├── model_converters └── publish_model.py ├── slurm_test.sh ├── slurm_train.sh ├── test.py └── train.py /.gitattributes: -------------------------------------------------------------------------------- 1 | # Auto detect text files and perform LF normalization 2 | * text=auto 3 | -------------------------------------------------------------------------------- /.github/CONTRIBUTING.md: -------------------------------------------------------------------------------- 1 | We appreciate all contributions to improve MMRotate. Please refer to [CONTRIBUTING.md](https://github.com/open-mmlab/mmcv/blob/master/CONTRIBUTING.md) in MMCV for more details about the contributing guideline. 2 | -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/2-feature-request.yml: -------------------------------------------------------------------------------- 1 | name: 🚀 Feature request 2 | description: Suggest an idea for this project 3 | labels: "kind/enhancement,status/unconfirmed" 4 | title: "[Feature] " 5 | 6 | body: 7 | - type: markdown 8 | attributes: 9 | value: | 10 | We strongly appreciate you creating a PR to implement this feature [here](https://github.com/open-mmlab/mmrotate/pulls)! 11 | If you need our help, please fill in as much of the following form as you're able to. 12 | 13 | **The less clear the description, the longer it will take to solve it.** 14 | 15 | - type: textarea 16 | attributes: 17 | label: What's the feature? 18 | description: | 19 | Tell us more about the feature and how this feature can help. 20 | placeholder: | 21 | E.g., It is inconvenient when \[....\]. 22 | This feature can \[....\]. 23 | validations: 24 | required: true 25 | 26 | - type: textarea 27 | attributes: 28 | label: Any other context? 29 | description: | 30 | Have you considered any alternative solutions or features? If so, what are they? 31 | Also, feel free to add any other context or screenshots about the feature request here. 32 | -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/3-new-model.yml: -------------------------------------------------------------------------------- 1 | name: "\U0001F31F New model/dataset/scheduler addition" 2 | description: Submit a proposal/request to implement a new model / dataset / scheduler 3 | labels: "kind/feature,status/unconfirmed" 4 | title: "[New Models] " 5 | 6 | 7 | body: 8 | - type: textarea 9 | id: description-request 10 | validations: 11 | required: true 12 | attributes: 13 | label: Model/Dataset/Scheduler description 14 | description: | 15 | Put any and all important information relative to the model/dataset/scheduler 16 | 17 | - type: checkboxes 18 | attributes: 19 | label: Open source status 20 | description: | 21 | Please provide the open-source status, which would be very helpful 22 | options: 23 | - label: "The model implementation is available" 24 | - label: "The model weights are available." 25 | 26 | - type: textarea 27 | id: additional-info 28 | attributes: 29 | label: Provide useful links for the implementation 30 | description: | 31 | Please provide information regarding the implementation, the weights, and the authors. 32 | Please mention the authors by @gh-username if you're aware of their usernames. 33 | -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/4-documentation.yml: -------------------------------------------------------------------------------- 1 | name: 📚 Documentation 2 | description: Report an issue related to the documentation. 3 | labels: "kind/doc,status/unconfirmed" 4 | title: "[Docs] " 5 | 6 | body: 7 | - type: dropdown 8 | id: branch 9 | attributes: 10 | label: Branch 11 | description: This issue is related to the 12 | options: 13 | - master branch https://mmrotate.readthedocs.io/en/latest/ 14 | - 1.x branch https://mmrotate.readthedocs.io/en/1.x/ 15 | validations: 16 | required: true 17 | 18 | - type: textarea 19 | attributes: 20 | label: 📚 The doc issue 21 | description: > 22 | A clear and concise description the issue. 23 | validations: 24 | required: true 25 | 26 | - type: textarea 27 | attributes: 28 | label: Suggest a potential alternative/fix 29 | description: > 30 | Tell us how we could improve the documentation in this regard. 31 | - type: markdown 32 | attributes: 33 | value: > 34 | Thanks for contributing 🎉! 35 | -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/config.yml: -------------------------------------------------------------------------------- 1 | blank_issues_enabled: false 2 | 3 | contact_links: 4 | - name: Common Issues 5 | url: https://mmrotate.readthedocs.io/en/latest/faq.html 6 | about: Check if your issue already has solutions 7 | - name: MMRotate Documentation 8 | url: https://mmrotate.readthedocs.io/en/latest/ 9 | about: Check if your question is answered in docs 10 | -------------------------------------------------------------------------------- /.github/pull_request_template.md: -------------------------------------------------------------------------------- 1 | Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers. 2 | 3 | ## Motivation 4 | 5 | Please describe the motivation of this PR and the goal you want to achieve through this PR. 6 | 7 | ## Modification 8 | 9 | Please briefly describe what modification is made in this PR. 10 | 11 | ## BC-breaking (Optional) 12 | 13 | Does the modification introduce changes that break the back-compatibility of the downstream repos? 14 | If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR. 15 | 16 | ## Use cases (Optional) 17 | 18 | If this PR introduces a new feature, it is better to list some use cases here, and update the documentation. 19 | 20 | ## Checklist 21 | 22 | 1. Pre-commit or other linting tools are used to fix the potential lint issues. 23 | 2. The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness. 24 | 3. The documentation has been modified accordingly, like docstring or example tutorials. 25 | -------------------------------------------------------------------------------- /.github/workflows/lint.yml: -------------------------------------------------------------------------------- 1 | name: lint 2 | 3 | on: [push, pull_request] 4 | 5 | concurrency: 6 | group: ${{ github.workflow }}-${{ github.ref }} 7 | cancel-in-progress: true 8 | 9 | jobs: 10 | lint: 11 | runs-on: ubuntu-latest 12 | steps: 13 | - uses: actions/checkout@v2 14 | - name: Set up Python 3.7 15 | uses: actions/setup-python@v2 16 | with: 17 | python-version: 3.7 18 | - name: Install pre-commit hook 19 | run: | 20 | pip install pre-commit 21 | pre-commit install 22 | - name: Linting 23 | run: pre-commit run --all-files 24 | - name: Check docstring coverage 25 | run: | 26 | pip install interrogate 27 | interrogate -v --ignore-init-method --ignore-module --ignore-nested-functions --ignore-regex "__repr__" --fail-under 90 mmrotate 28 | -------------------------------------------------------------------------------- /.github/workflows/publish-to-pypi.yml: -------------------------------------------------------------------------------- 1 | name: deploy 2 | 3 | on: push 4 | 5 | concurrency: 6 | group: ${{ github.workflow }}-${{ github.ref }} 7 | cancel-in-progress: true 8 | 9 | jobs: 10 | build-n-publish: 11 | runs-on: ubuntu-latest 12 | if: startsWith(github.event.ref, 'refs/tags') 13 | steps: 14 | - uses: actions/checkout@v2 15 | - name: Set up Python 3.7 16 | uses: actions/setup-python@v1 17 | with: 18 | python-version: 3.7 19 | - name: Build MMRotate 20 | run: | 21 | pip install wheel 22 | python setup.py sdist bdist_wheel 23 | - name: Publish distribution to PyPI 24 | run: | 25 | pip install twine 26 | twine upload dist/* -u __token__ -p ${{ secrets.pypi_password }} 27 | -------------------------------------------------------------------------------- /.github/workflows/scripts/get_mmcv_var.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | 3 | TORCH=$1 4 | CUDA=$2 5 | 6 | # 10.2 -> cu102 7 | MMCV_CUDA="cu`echo ${CUDA} | tr -d '.'`" 8 | 9 | # MMCV only provides pre-compiled packages for torch 1.x.0 10 | # which works for any subversions of torch 1.x. 11 | # We force the torch version to be 1.x.0 to ease package searching 12 | # and avoid unnecessary rebuild during MMCV's installation. 13 | TORCH_VER_ARR=(${TORCH//./ }) 14 | TORCH_VER_ARR[2]=0 15 | printf -v MMCV_TORCH "%s." "${TORCH_VER_ARR[@]}" 16 | MMCV_TORCH=${MMCV_TORCH%?} # Remove the last dot 17 | 18 | echo "MMCV_CUDA=${MMCV_CUDA}" >> $GITHUB_ENV 19 | echo "MMCV_TORCH=${MMCV_TORCH}" >> $GITHUB_ENV 20 | -------------------------------------------------------------------------------- /.github/workflows/test_mim.yml: -------------------------------------------------------------------------------- 1 | name: test-mim 2 | 3 | on: 4 | push: 5 | paths: 6 | - 'model-index.yml' 7 | - 'configs/**' 8 | 9 | pull_request: 10 | paths: 11 | - 'model-index.yml' 12 | - 'configs/**' 13 | 14 | concurrency: 15 | group: ${{ github.workflow }}-${{ github.ref }} 16 | cancel-in-progress: true 17 | 18 | jobs: 19 | build_cpu: 20 | runs-on: ubuntu-18.04 21 | strategy: 22 | matrix: 23 | python-version: [3.7] 24 | torch: [1.8.0] 25 | include: 26 | - torch: 1.8.0 27 | torch_version: torch1.8 28 | torchvision: 0.9.0 29 | steps: 30 | - uses: actions/checkout@v2 31 | - name: Set up Python ${{ matrix.python-version }} 32 | uses: actions/setup-python@v2 33 | with: 34 | python-version: ${{ matrix.python-version }} 35 | - name: Upgrade pip 36 | run: pip install pip --upgrade 37 | - name: Install PyTorch 38 | run: pip install torch==${{matrix.torch}}+cpu torchvision==${{matrix.torchvision}}+cpu -f https://download.pytorch.org/whl/torch_stable.html 39 | - name: Install openmim 40 | run: pip install openmim 41 | - name: Build and install 42 | run: rm -rf .eggs && mim install -e . 43 | - name: test commands of mim 44 | run: mim search mmrotate 45 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | *.py[cod] 4 | *$py.class 5 | 6 | # C extensions 7 | *.so 8 | 9 | # Distribution / packaging 10 | .Python 11 | build/ 12 | develop-eggs/ 13 | dist/ 14 | downloads/ 15 | eggs/ 16 | .eggs/ 17 | lib/ 18 | lib64/ 19 | parts/ 20 | sdist/ 21 | var/ 22 | wheels/ 23 | *.egg-info/ 24 | .installed.cfg 25 | *.egg 26 | MANIFEST 27 | 28 | # PyInstaller 29 | # Usually these files are written by a python script from a template 30 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 31 | *.manifest 32 | *.spec 33 | 34 | # Installer logs 35 | pip-log.txt 36 | pip-delete-this-directory.txt 37 | 38 | # Unit test / coverage reports 39 | htmlcov/ 40 | .tox/ 41 | .nox/ 42 | .coverage 43 | .coverage.* 44 | .cache 45 | nosetests.xml 46 | coverage.xml 47 | *.cover 48 | .hypothesis/ 49 | .pytest_cache/ 50 | .idea/* 51 | 52 | # Translations 53 | *.mo 54 | *.pot 55 | 56 | # Django stuff: 57 | *.log 58 | local_settings.py 59 | db.sqlite3 60 | 61 | # Flask stuff: 62 | instance/ 63 | .webassets-cache 64 | 65 | # Scrapy stuff: 66 | .scrapy 67 | 68 | # Sphinx documentation 69 | docs/en/_build/ 70 | docs/zh_cn/_build/ 71 | src 72 | 73 | # PyBuilder 74 | target/ 75 | 76 | # Jupyter Notebook 77 | .ipynb_checkpoints 78 | 79 | # IPython 80 | profile_default/ 81 | ipython_config.py 82 | 83 | # pyenv 84 | .python-version 85 | 86 | # celery beat schedule file 87 | celerybeat-schedule 88 | 89 | # SageMath parsed files 90 | *.sage.py 91 | 92 | # Environments 93 | .env 94 | .venv 95 | env/ 96 | venv/ 97 | ENV/ 98 | env.bak/ 99 | venv.bak/ 100 | 101 | # Spyder project settings 102 | .spyderproject 103 | .spyproject 104 | 105 | # Rope project settings 106 | .ropeproject 107 | 108 | # mkdocs documentation 109 | /site 110 | 111 | # mypy 112 | .mypy_cache/ 113 | .dmypy.json 114 | dmypy.json 115 | 116 | # Pyre type checker 117 | .pyre/ 118 | .DS_Store 119 | .idea 120 | *work_dirs* 121 | data/ 122 | data 123 | tmp 124 | debug_demo* 125 | 126 | # custom 127 | mmrotate/.mim 128 | .bak 129 | .vscode 130 | outputs/ 131 | checkpoints* -------------------------------------------------------------------------------- /.pre-commit-config-zh-cn.yaml: -------------------------------------------------------------------------------- 1 | exclude: ^tests/data/ 2 | repos: 3 | - repo: https://gitee.com/openmmlab/mirrors-flake8 4 | rev: 5.0.4 5 | hooks: 6 | - id: flake8 7 | - repo: https://gitee.com/openmmlab/mirrors-isort 8 | rev: 5.10.1 9 | hooks: 10 | - id: isort 11 | - repo: https://gitee.com/openmmlab/mirrors-yapf 12 | rev: v0.32.0 13 | hooks: 14 | - id: yapf 15 | - repo: https://gitee.com/openmmlab/mirrors-pre-commit-hooks 16 | rev: v4.3.0 17 | hooks: 18 | - id: trailing-whitespace 19 | - id: check-yaml 20 | - id: end-of-file-fixer 21 | - id: requirements-txt-fixer 22 | - id: double-quote-string-fixer 23 | - id: check-merge-conflict 24 | - id: fix-encoding-pragma 25 | args: ["--remove"] 26 | - id: mixed-line-ending 27 | args: ["--fix=lf"] 28 | - repo: https://gitee.com/openmmlab/mirrors-mdformat 29 | rev: 0.7.9 30 | hooks: 31 | - id: mdformat 32 | args: ["--number"] 33 | additional_dependencies: 34 | - mdformat-openmmlab 35 | - mdformat_frontmatter 36 | - linkify-it-py 37 | - repo: https://gitee.com/openmmlab/mirrors-codespell 38 | rev: v2.2.1 39 | hooks: 40 | - id: codespell 41 | - repo: https://gitee.com/openmmlab/mirrors-docformatter 42 | rev: v1.3.1 43 | hooks: 44 | - id: docformatter 45 | args: ["--in-place", "--wrap-descriptions", "79"] 46 | - repo: https://gitee.com/openmmlab/mirrors-pyupgrade 47 | rev: v3.0.0 48 | hooks: 49 | - id: pyupgrade 50 | args: ["--py36-plus"] 51 | - repo: https://github.com/open-mmlab/pre-commit-hooks 52 | rev: v0.2.0 53 | hooks: 54 | - id: check-copyright 55 | args: ["mmyolo", "tests"] 56 | # - repo: https://gitee.com/openmmlab/mirrors-mypy 57 | # rev: v0.812 58 | # hooks: 59 | # - id: mypy 60 | # exclude: "docs" 61 | -------------------------------------------------------------------------------- /.pre-commit-config.yaml: -------------------------------------------------------------------------------- 1 | repos: 2 | - repo: https://github.com/PyCQA/flake8 3 | rev: 5.0.4 4 | hooks: 5 | - id: flake8 6 | - repo: https://github.com/asottile/seed-isort-config 7 | rev: v2.2.0 8 | hooks: 9 | - id: seed-isort-config 10 | - repo: https://github.com/timothycrosley/isort 11 | rev: 4.3.21 12 | hooks: 13 | - id: isort 14 | - repo: https://github.com/pre-commit/mirrors-yapf 15 | rev: v0.32.0 16 | hooks: 17 | - id: yapf 18 | - repo: https://github.com/pre-commit/pre-commit-hooks 19 | rev: v4.3.0 20 | hooks: 21 | - id: trailing-whitespace 22 | - id: check-yaml 23 | - id: end-of-file-fixer 24 | - id: requirements-txt-fixer 25 | - id: double-quote-string-fixer 26 | - id: check-merge-conflict 27 | - id: fix-encoding-pragma 28 | args: [ "--remove" ] 29 | - id: mixed-line-ending 30 | args: [ "--fix=lf" ] 31 | - repo: https://github.com/codespell-project/codespell 32 | rev: v2.2.1 33 | hooks: 34 | - id: codespell 35 | args: [ '--ignore-words-list', 'DOTA' ] 36 | - repo: https://github.com/executablebooks/mdformat 37 | rev: 0.7.9 38 | hooks: 39 | - id: mdformat 40 | args: ["--number"] 41 | additional_dependencies: 42 | - mdformat-openmmlab 43 | - mdformat_frontmatter 44 | - linkify-it-py 45 | - repo: https://github.com/myint/docformatter 46 | rev: v1.3.1 47 | hooks: 48 | - id: docformatter 49 | args: [ "--in-place", "--wrap-descriptions", "79" ] 50 | - repo: https://github.com/open-mmlab/pre-commit-hooks 51 | rev: v0.2.0 52 | hooks: 53 | - id: check-copyright 54 | args: ["mmrotate", "tests", "demo", "tools"] 55 | -------------------------------------------------------------------------------- /.readthedocs.yml: -------------------------------------------------------------------------------- 1 | version: 2 2 | 3 | formats: all 4 | 5 | python: 6 | version: 3.7 7 | install: 8 | - requirements: requirements/docs.txt 9 | - requirements: requirements/readthedocs.txt 10 | -------------------------------------------------------------------------------- /CITATION.cff: -------------------------------------------------------------------------------- 1 | cff-version: 1.2.0 2 | message: "If you use this software, please cite it as below." 3 | authors: 4 | - name: "MMRotate Contributors" 5 | title: "OpenMMLab rotated object detection toolbox and benchmark" 6 | date-released: 2022-02-18 7 | url: "https://github.com/open-mmlab/mmrotate" 8 | license: Apache-2.0 9 | -------------------------------------------------------------------------------- /MANIFEST.in: -------------------------------------------------------------------------------- 1 | include requirements/*.txt 2 | include mmrotate/.mim/model-index.yml 3 | recursive-include mmrotate/.mim/configs *.py *.yml 4 | recursive-include mmrotate/.mim/tools *.py *.sh 5 | -------------------------------------------------------------------------------- /configs/_base_/default_runtime.py: -------------------------------------------------------------------------------- 1 | default_scope = 'mmrotate' 2 | 3 | default_hooks = dict( 4 | timer=dict(type='IterTimerHook'), 5 | logger=dict(type='LoggerHook', interval=50), 6 | param_scheduler=dict(type='ParamSchedulerHook'), 7 | checkpoint=dict(type='CheckpointHook', interval=1), 8 | sampler_seed=dict(type='DistSamplerSeedHook'), 9 | visualization=dict(type='mmdet.DetVisualizationHook')) 10 | 11 | env_cfg = dict( 12 | cudnn_benchmark=False, 13 | mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0), 14 | dist_cfg=dict(backend='nccl'), 15 | ) 16 | 17 | vis_backends = [dict(type='LocalVisBackend')] 18 | visualizer = dict( 19 | type='RotLocalVisualizer', vis_backends=vis_backends, name='visualizer') 20 | log_processor = dict(type='LogProcessor', window_size=50, by_epoch=True) 21 | 22 | log_level = 'INFO' 23 | load_from = None 24 | resume = False 25 | -------------------------------------------------------------------------------- /configs/_base_/schedules/schedule_1x.py: -------------------------------------------------------------------------------- 1 | # training schedule for 1x 2 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=12, val_interval=1) 3 | val_cfg = dict(type='ValLoop') 4 | test_cfg = dict(type='TestLoop') 5 | 6 | # learning rate 7 | param_scheduler = [ 8 | dict( 9 | type='LinearLR', 10 | start_factor=1.0 / 3, 11 | by_epoch=False, 12 | begin=0, 13 | end=500), 14 | dict( 15 | type='MultiStepLR', 16 | begin=0, 17 | end=12, 18 | by_epoch=True, 19 | milestones=[8, 11], 20 | gamma=0.1) 21 | ] 22 | 23 | # optimizer 24 | optim_wrapper = dict( 25 | type='OptimWrapper', 26 | optimizer=dict(type='SGD', lr=0.0025, momentum=0.9, weight_decay=0.0001), 27 | clip_grad=dict(max_norm=35, norm_type=2)) 28 | -------------------------------------------------------------------------------- /configs/_base_/schedules/schedule_3x.py: -------------------------------------------------------------------------------- 1 | # training schedule for 1x 2 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=36, val_interval=1) 3 | val_cfg = dict(type='ValLoop') 4 | test_cfg = dict(type='TestLoop') 5 | 6 | # learning rate 7 | param_scheduler = [ 8 | dict( 9 | type='LinearLR', 10 | start_factor=1.0 / 3, 11 | by_epoch=False, 12 | begin=0, 13 | end=500), 14 | dict( 15 | type='MultiStepLR', 16 | begin=0, 17 | end=36, 18 | by_epoch=True, 19 | milestones=[24, 33], 20 | gamma=0.1) 21 | ] 22 | 23 | # optimizer 24 | optim_wrapper = dict( 25 | type='OptimWrapper', 26 | optimizer=dict(type='SGD', lr=0.0025, momentum=0.9, weight_decay=0.0001), 27 | clip_grad=dict(max_norm=35, norm_type=2)) 28 | -------------------------------------------------------------------------------- /configs/_base_/schedules/schedule_40e.py: -------------------------------------------------------------------------------- 1 | # training schedule for 1x 2 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=40, val_interval=1) 3 | val_cfg = dict(type='ValLoop') 4 | test_cfg = dict(type='TestLoop') 5 | 6 | # learning rate 7 | param_scheduler = [ 8 | dict( 9 | type='LinearLR', 10 | start_factor=1.0 / 3, 11 | by_epoch=False, 12 | begin=0, 13 | end=500), 14 | dict( 15 | type='MultiStepLR', 16 | begin=0, 17 | end=40, 18 | by_epoch=True, 19 | milestones=[24, 32, 38], 20 | gamma=0.1) 21 | ] 22 | 23 | # optimizer 24 | optim_wrapper = dict( 25 | type='OptimWrapper', 26 | optimizer=dict(type='SGD', lr=0.0025, momentum=0.9, weight_decay=0.0001), 27 | clip_grad=dict(max_norm=35, norm_type=2)) 28 | -------------------------------------------------------------------------------- /configs/_base_/schedules/schedule_6x.py: -------------------------------------------------------------------------------- 1 | # training schedule for 1x 2 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=72, val_interval=1) 3 | val_cfg = dict(type='ValLoop') 4 | test_cfg = dict(type='TestLoop') 5 | 6 | # learning rate 7 | param_scheduler = [ 8 | dict( 9 | type='LinearLR', 10 | start_factor=1.0 / 3, 11 | by_epoch=False, 12 | begin=0, 13 | end=500), 14 | dict( 15 | type='MultiStepLR', 16 | begin=0, 17 | end=72, 18 | by_epoch=True, 19 | milestones=[48, 66], 20 | gamma=0.1) 21 | ] 22 | 23 | # optimizer 24 | optim_wrapper = dict( 25 | type='OptimWrapper', 26 | optimizer=dict(type='SGD', lr=0.0025, momentum=0.9, weight_decay=0.0001), 27 | clip_grad=dict(max_norm=35, norm_type=2)) 28 | -------------------------------------------------------------------------------- /configs/cfa/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: cfa 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://openaccess.thecvf.com/content/CVPR2021/papers/Guo_Beyond_Bounding-Box_Convex-Hull_Feature_Adaptation_for_Oriented_and_Densely_Packed_CVPR_2021_paper.pdf 13 | Title: 'Beyond Bounding-Box: Convex-hull Feature Adaptation for Oriented and Densely Packed Object Detection' 14 | README: configs/cfa/README.md 15 | 16 | Models: 17 | - Name: cfa-qbox_r50_fpn_1x_dota 18 | In Collection: cfa 19 | Config: configs/cfa/cfa-qbox_r50_fpn_1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 69.63 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/cfa/cfa_r50_fpn_1x_dota_le135/cfa_r50_fpn_1x_dota_le135-aed1cbc6.pth 28 | 29 | - Name: cfa-qbox_r50_fpn_40e_dota 30 | In Collection: cfa 31 | Config: configs/cfa/cfa-qbox_r50_fpn_40e_dota.py 32 | Metadata: 33 | Training Data: DOTAv1.0 34 | Results: 35 | - Task: Oriented Object Detection 36 | Dataset: DOTAv1.0 37 | Metrics: 38 | mAP: 73.45 39 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/cfa/cfa_r50_fpn_40e_dota_oc/cfa_r50_fpn_40e_dota_oc-2f387232.pth 40 | -------------------------------------------------------------------------------- /configs/convnext/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: convnext 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - AdamW 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 3090 9 | Architecture: 10 | - ConvNeXt 11 | Paper: 12 | URL: https://arxiv.org/abs/2201.03545 13 | Title: 'A ConvNet for the 2020s' 14 | README: configs/convnext/README.md 15 | 16 | Models: 17 | - Name: rotated-retinanet-rbox-le90_convnext-tiny_fpn_kld-stable_adamw-1x_dota 18 | In Collection: convnext 19 | Config: configs/convnext/rotated-retinanet-rbox-le90_convnext-tiny_fpn_kld-stable_adamw-1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 74.49 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/convnext/rotated_retinanet_obb_kld_stable_convnext_adamw_fpn_1x_dota_le90/rotated_retinanet_obb_kld_stable_convnext_adamw_fpn_1x_dota_le90-388184f6.pth 28 | -------------------------------------------------------------------------------- /configs/convnext/rotated-retinanet-rbox-le90_convnext-tiny_fpn_kld-stable_adamw-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = [ 2 | '../kld/rotated-retinanet-rbox-le90_r50_fpn_kld-stable_adamw-1x_dota.py' 3 | ] 4 | 5 | # please install mmcls>=1.0.0rc0 6 | # import mmcls.models to trigger register_module in mmcls 7 | custom_imports = dict(imports=['mmcls.models'], allow_failed_imports=False) 8 | checkpoint_file = 'https://download.openmmlab.com/mmclassification/v0/convnext/downstream/convnext-tiny_3rdparty_32xb128-noema_in1k_20220301-795e9634.pth' # noqa 9 | 10 | model = dict( 11 | backbone=dict( 12 | _delete_=True, 13 | type='mmcls.ConvNeXt', 14 | arch='tiny', 15 | out_indices=[0, 1, 2, 3], 16 | drop_path_rate=0.4, 17 | layer_scale_init_value=1.0, 18 | gap_before_final_norm=False, 19 | init_cfg=dict( 20 | type='Pretrained', checkpoint=checkpoint_file, 21 | prefix='backbone.')), 22 | neck=dict(in_channels=[96, 192, 384, 768])) 23 | 24 | lr_config = dict(warmup_iters=1000) 25 | find_unused_parameters = True 26 | -------------------------------------------------------------------------------- /configs/csl/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: csl 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x Quadro RTX 8000 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://link.springer.com/chapter/10.1007/978-3-030-58598-3_40 13 | Title: 'Arbitrary-Oriented Object Detection with Circular Smooth Label' 14 | README: configs/csl/README.md 15 | 16 | Models: 17 | - Name: rotated-retinanet-rbox-le90_r50_fpn_csl-gaussian_amp-1x_dota 18 | In Collection: csl 19 | Config: configs/csl/rotated-retinanet-rbox-le90_r50_fpn_csl-gaussian_amp-1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 69.51 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/csl/rotated_retinanet_obb_csl_gaussian_r50_fpn_fp16_1x_dota_le90/rotated_retinanet_obb_csl_gaussian_r50_fpn_fp16_1x_dota_le90-b4271aed.pth 28 | -------------------------------------------------------------------------------- /configs/csl/rotated-retinanet-rbox-le90_r50_fpn_csl-gaussian_amp-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = \ 2 | ['../rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_amp-1x_dota.py'] 3 | 4 | angle_version = 'le90' 5 | model = dict( 6 | bbox_head=dict( 7 | type='AngleBranchRetinaHead', 8 | angle_coder=dict( 9 | type='CSLCoder', 10 | angle_version=angle_version, 11 | omega=4, 12 | window='gaussian', 13 | radius=3), 14 | loss_angle=dict( 15 | type='SmoothFocalLoss', gamma=2.0, alpha=0.25, loss_weight=0.8))) 16 | -------------------------------------------------------------------------------- /configs/gliding_vertex/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: gliding_vertex 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://arxiv.org/pdf/1911.09358.pdf 13 | Title: 'Gliding Vertex on the Horizontal Bounding Box for Multi-Oriented Object Detection' 14 | README: configs/cfa/README.md 15 | 16 | Models: 17 | - Name: gliding-vertex-rbox_r50_fpn_1x_dota 18 | In Collection: gliding_vertex 19 | Config: configs/gliding_vertex/gliding-vertex-rbox_r50_fpn_1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 73.23 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/gliding_vertex/gliding_vertex_r50_fpn_1x_dota_le90/gliding_vertex_r50_fpn_1x_dota_le90-12e7423c.pth 28 | -------------------------------------------------------------------------------- /configs/gwd/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: gwd 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://arxiv.org/pdf/2101.11952.pdf 13 | Title: 'Rethinking Rotated Object Detection with Gaussian Wasserstein Distance Loss' 14 | README: configs/gwd/README.md 15 | 16 | Models: 17 | - Name: rotated-retinanet-hbox-oc_r50_fpn_gwd_1x_dota 18 | In Collection: gwd 19 | Config: configs/gwd/rotated-retinanet-hbox-oc_r50_fpn_gwd_1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 69.55 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/gwd/rotated_retinanet_hbb_gwd_r50_fpn_1x_dota_oc/rotated_retinanet_hbb_gwd_r50_fpn_1x_dota_oc-41fd7805.pth 28 | -------------------------------------------------------------------------------- /configs/gwd/rotated-retinanet-hbox-oc_r50_fpn_gwd_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-hbox-oc_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | reg_decoded_bbox=True, 6 | loss_bbox=dict(type='GDLoss', loss_type='gwd', loss_weight=5.0))) 7 | -------------------------------------------------------------------------------- /configs/gwd/rotated-retinanet-rbox-le135_r50_fpn_gwd_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-obb-le135_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | reg_decoded_bbox=True, 6 | loss_bbox=dict(type='GDLoss', loss_type='gwd', loss_weight=5.0))) 7 | -------------------------------------------------------------------------------- /configs/gwd/rotated-retinanet-rbox-le90_r50_fpn_gwd_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = [ 2 | '../rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py' 3 | ] 4 | 5 | model = dict( 6 | bbox_head=dict( 7 | reg_decoded_bbox=True, 8 | loss_bbox=dict(type='GDLoss', loss_type='gwd', loss_weight=5.0))) 9 | -------------------------------------------------------------------------------- /configs/h2rbox/dotav15/h2rbox-le90_r50_fpn_adamw-3x_dotav15.py: -------------------------------------------------------------------------------- 1 | _base_ = './h2rbox-le90_r50_fpn_adamw-1x_dotav15.py' 2 | 3 | # training schedule for 3x 4 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=36, val_interval=12) 5 | 6 | param_scheduler = [ 7 | dict( 8 | type='LinearLR', 9 | start_factor=1.0 / 3, 10 | by_epoch=False, 11 | begin=0, 12 | end=500), 13 | dict( 14 | type='MultiStepLR', 15 | begin=0, 16 | end=36, 17 | by_epoch=True, 18 | milestones=[24, 33], 19 | gamma=0.1) 20 | ] 21 | -------------------------------------------------------------------------------- /configs/h2rbox/dotav2/h2rbox-le90_r50_fpn_adamw-3x_dotav2.py: -------------------------------------------------------------------------------- 1 | _base_ = './h2rbox-le90_r50_fpn_adamw-1x_dotav2.py' 2 | 3 | # training schedule for 3x 4 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=36, val_interval=12) 5 | 6 | param_scheduler = [ 7 | dict( 8 | type='LinearLR', 9 | start_factor=1.0 / 3, 10 | by_epoch=False, 11 | begin=0, 12 | end=500), 13 | dict( 14 | type='MultiStepLR', 15 | begin=0, 16 | end=36, 17 | by_epoch=True, 18 | milestones=[24, 33], 19 | gamma=0.1) 20 | ] 21 | -------------------------------------------------------------------------------- /configs/h2rbox/h2rbox-le90_r50_fpn_adamw-1x_dota-ms.py: -------------------------------------------------------------------------------- 1 | _base_ = './h2rbox-le90_r50_fpn_adamw-1x_dota.py' 2 | data_root = '/data/nas/dataset_share/DOTA/split_ms_dota1_0/' 3 | 4 | train_dataloader = dict(dataset=dict(data_root=data_root)) 5 | val_dataloader = dict(dataset=dict(data_root=data_root)) 6 | test_dataloader = dict(dataset=dict(data_root=data_root)) 7 | -------------------------------------------------------------------------------- /configs/h2rbox/h2rbox-le90_r50_fpn_adamw-3x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './h2rbox-le90_r50_fpn_adamw-1x_dota.py' 2 | 3 | # training schedule for 3x 4 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=36, val_interval=12) 5 | 6 | param_scheduler = [ 7 | dict( 8 | type='LinearLR', 9 | start_factor=1.0 / 3, 10 | by_epoch=False, 11 | begin=0, 12 | end=500), 13 | dict( 14 | type='MultiStepLR', 15 | begin=0, 16 | end=36, 17 | by_epoch=True, 18 | milestones=[24, 33], 19 | gamma=0.1) 20 | ] 21 | -------------------------------------------------------------------------------- /configs/h2rbox_v2/h2rbox_v2-le90_r50_fpn_ms_rr-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = ['h2rbox_v2-le90_r50_fpn-1x_dota.py'] 2 | 3 | # load hbox annotations 4 | train_pipeline = [ 5 | dict(type='mmdet.LoadImageFromFile', backend_args={{_base_.backend_args}}), 6 | dict(type='mmdet.LoadAnnotations', with_bbox=True, box_type='qbox'), 7 | # Horizontal GTBox, (x1,y1,x2,y2) 8 | dict(type='ConvertBoxType', box_type_mapping=dict(gt_bboxes='hbox')), 9 | # Horizontal GTBox, (x,y,w,h,theta) 10 | dict(type='ConvertBoxType', box_type_mapping=dict(gt_bboxes='rbox')), 11 | dict(type='mmdet.Resize', scale=(1024, 1024), keep_ratio=True), 12 | dict( 13 | type='mmdet.RandomFlip', 14 | prob=0.75, 15 | direction=['horizontal', 'vertical', 'diagonal']), 16 | dict(type='RandomRotate', prob=1, angle_range=180), 17 | dict(type='mmdet.PackDetInputs') 18 | ] 19 | 20 | train_dataloader = dict(dataset=dict()) 21 | 22 | data_root = 'data/split_ms_dota/' 23 | 24 | train_dataloader = dict( 25 | dataset=dict(data_root=data_root, pipeline=train_pipeline)) 26 | val_dataloader = dict(dataset=dict(data_root=data_root)) 27 | test_dataloader = dict(dataset=dict(data_root=data_root)) 28 | -------------------------------------------------------------------------------- /configs/h2rbox_v2/h2rbox_v2-le90_r50_fpn_rr-6x_hrsc.py: -------------------------------------------------------------------------------- 1 | _base_ = ['h2rbox_v2-le90_r50_fpn-6x_hrsc.py'] 2 | 3 | # load hbox annotations 4 | train_pipeline = [ 5 | dict(type='mmdet.LoadImageFromFile', backend_args={{_base_.backend_args}}), 6 | dict(type='mmdet.LoadAnnotations', with_bbox=True, box_type='qbox'), 7 | dict(type='mmdet.FixShapeResize', width=800, height=800, keep_ratio=True), 8 | # Horizontal GTBox, (x1,y1,x2,y2) 9 | dict(type='ConvertBoxType', box_type_mapping=dict(gt_bboxes='hbox')), 10 | # Horizontal GTBox, (x,y,w,h,theta) 11 | dict(type='ConvertBoxType', box_type_mapping=dict(gt_bboxes='rbox')), 12 | dict( 13 | type='mmdet.RandomFlip', 14 | prob=0.75, 15 | direction=['horizontal', 'vertical', 'diagonal']), 16 | dict(type='RandomRotate', prob=1, angle_range=180), 17 | dict(type='mmdet.PackDetInputs') 18 | ] 19 | 20 | train_dataloader = dict(dataset=dict(pipeline=train_pipeline)) 21 | -------------------------------------------------------------------------------- /configs/kfiou/r3det-oc_r50_fpn_kfiou-ln_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = ['../r3det/r3det-oc_r50_fpn_1x_dota.py'] 2 | 3 | angle_version = 'oc' 4 | model = dict( 5 | bbox_head_init=dict( 6 | loss_bbox_type='kfiou', 7 | loss_bbox=dict(type='KFLoss', fun='ln', loss_weight=5.0)), 8 | bbox_head_refine=[ 9 | dict( 10 | type='R3RefineHead', 11 | num_classes=15, 12 | in_channels=256, 13 | stacked_convs=4, 14 | feat_channels=256, 15 | frm_cfg=dict( 16 | type='FRM', feat_channels=256, strides=[8, 16, 32, 64, 128]), 17 | anchor_generator=dict( 18 | type='PseudoRotatedAnchorGenerator', 19 | strides=[8, 16, 32, 64, 128]), 20 | bbox_coder=dict( 21 | type='DeltaXYWHTRBBoxCoder', 22 | angle_version=angle_version, 23 | norm_factor=None, 24 | edge_swap=False, 25 | proj_xy=False, 26 | target_means=(0.0, 0.0, 0.0, 0.0, 0.0), 27 | target_stds=(1.0, 1.0, 1.0, 1.0, 1.0)), 28 | loss_cls=dict( 29 | type='mmdet.FocalLoss', 30 | use_sigmoid=True, 31 | gamma=2.0, 32 | alpha=0.25, 33 | loss_weight=1.0), 34 | loss_bbox_type='kfiou', 35 | loss_bbox=dict(type='KFLoss', fun='ln', loss_weight=5.0)) 36 | ]) 37 | -------------------------------------------------------------------------------- /configs/kfiou/roi-trans-le90_r50_fpn_kfiou-ln_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../roi_trans/roi-trans-le90_r50_fpn_1x_dota.py' 2 | 3 | angle_version = 'le90' 4 | model = dict( 5 | roi_head=dict(bbox_head=[ 6 | dict( 7 | type='RotatedShared2FCBBoxHead', 8 | predict_box_type='rbox', 9 | in_channels=256, 10 | fc_out_channels=1024, 11 | roi_feat_size=7, 12 | num_classes=15, 13 | reg_predictor_cfg=dict(type='mmdet.Linear'), 14 | cls_predictor_cfg=dict(type='mmdet.Linear'), 15 | bbox_coder=dict( 16 | type='DeltaXYWHTHBBoxCoder', 17 | angle_version=angle_version, 18 | norm_factor=2, 19 | edge_swap=True, 20 | target_means=(.0, .0, .0, .0, .0), 21 | target_stds=(0.1, 0.1, 0.2, 0.2, 0.1), 22 | use_box_type=True), 23 | reg_class_agnostic=True, 24 | loss_cls=dict( 25 | type='mmdet.CrossEntropyLoss', 26 | use_sigmoid=False, 27 | loss_weight=1.0), 28 | loss_bbox_type='kfiou', 29 | loss_bbox=dict(type='KFLoss', fun='ln', loss_weight=5.0)), 30 | dict( 31 | type='RotatedShared2FCBBoxHead', 32 | predict_box_type='rbox', 33 | in_channels=256, 34 | fc_out_channels=1024, 35 | roi_feat_size=7, 36 | num_classes=15, 37 | reg_predictor_cfg=dict(type='mmdet.Linear'), 38 | cls_predictor_cfg=dict(type='mmdet.Linear'), 39 | bbox_coder=dict( 40 | type='DeltaXYWHTRBBoxCoder', 41 | angle_version=angle_version, 42 | norm_factor=None, 43 | edge_swap=True, 44 | proj_xy=True, 45 | target_means=[0., 0., 0., 0., 0.], 46 | target_stds=[0.05, 0.05, 0.1, 0.1, 0.05]), 47 | reg_class_agnostic=False, 48 | loss_cls=dict( 49 | type='mmdet.CrossEntropyLoss', 50 | use_sigmoid=False, 51 | loss_weight=1.0), 52 | loss_bbox_type='kfiou', 53 | loss_bbox=dict(type='KFLoss', fun='ln', loss_weight=5.0)) 54 | ])) 55 | -------------------------------------------------------------------------------- /configs/kfiou/rotated-retinanet-hbox-le135_r50_fpn_kfiou_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-rbox-le135_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | type='RotatedRetinaHead', 6 | loss_bbox_type='kfiou', 7 | loss_bbox=dict(type='KFLoss', loss_weight=5.0)), 8 | train_cfg=dict( 9 | assigner=dict(iou_calculator=dict(type='FakeRBboxOverlaps2D')))) 10 | -------------------------------------------------------------------------------- /configs/kfiou/rotated-retinanet-hbox-le90_r50_fpn_kfiou_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | type='RotatedRetinaHead', 6 | loss_bbox_type='kfiou', 7 | loss_bbox=dict(type='KFLoss', loss_weight=5.0)), 8 | train_cfg=dict( 9 | assigner=dict(iou_calculator=dict(type='FakeRBboxOverlaps2D')))) 10 | -------------------------------------------------------------------------------- /configs/kfiou/rotated-retinanet-hbox-oc_r50_fpn_kfiou_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-rbox-oc_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | type='RotatedRetinaHead', 6 | loss_bbox_type='kfiou', 7 | loss_bbox=dict(type='KFLoss', loss_weight=5.0)), 8 | train_cfg=dict( 9 | assigner=dict(iou_calculator=dict(type='FakeRBboxOverlaps2D')))) 10 | -------------------------------------------------------------------------------- /configs/kfiou/s2anet-le135_r50_fpn_kfiou-ln_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../s2anet/s2anet-le135_r50_fpn_1x_dota.py' 2 | 3 | angle_version = 'le135' 4 | model = dict( 5 | bbox_head_init=dict( 6 | loss_bbox_type='kfiou', 7 | loss_bbox=dict(type='KFLoss', fun='ln', loss_weight=5.0)), 8 | bbox_head_refine=[ 9 | dict( 10 | type='S2ARefineHead', 11 | num_classes=15, 12 | in_channels=256, 13 | stacked_convs=2, 14 | feat_channels=256, 15 | frm_cfg=dict( 16 | type='AlignConv', 17 | feat_channels=256, 18 | kernel_size=3, 19 | strides=[8, 16, 32, 64, 128]), 20 | anchor_generator=dict( 21 | type='PseudoRotatedAnchorGenerator', 22 | strides=[8, 16, 32, 64, 128]), 23 | bbox_coder=dict( 24 | type='DeltaXYWHTRBBoxCoder', 25 | angle_version=angle_version, 26 | norm_factor=1, 27 | edge_swap=False, 28 | proj_xy=True, 29 | target_means=(0.0, 0.0, 0.0, 0.0, 0.0), 30 | target_stds=(1.0, 1.0, 1.0, 1.0, 1.0)), 31 | loss_cls=dict( 32 | type='mmdet.FocalLoss', 33 | use_sigmoid=True, 34 | gamma=2.0, 35 | alpha=0.25, 36 | loss_weight=1.0), 37 | loss_bbox_type='kfiou', 38 | loss_bbox=dict(type='KFLoss', fun='ln', loss_weight=5.0)) 39 | ]) 40 | -------------------------------------------------------------------------------- /configs/kld/r3det-oc_r50_fpn_kld-stable_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = ['../r3det/r3det-oc_r50_fpn_1x_dota.py'] 2 | 3 | angle_version = 'oc' 4 | model = dict( 5 | bbox_head_init=dict( 6 | reg_decoded_bbox=True, 7 | loss_bbox=dict( 8 | _delete_=True, 9 | type='GDLoss', 10 | loss_type='kld', 11 | fun='log1p', 12 | tau=1.0, 13 | sqrt=False, 14 | loss_weight=5.0)), 15 | bbox_head_refine=[ 16 | dict( 17 | type='R3RefineHead', 18 | num_classes=15, 19 | in_channels=256, 20 | stacked_convs=4, 21 | feat_channels=256, 22 | frm_cfg=dict( 23 | type='FRM', feat_channels=256, strides=[8, 16, 32, 64, 128]), 24 | anchor_generator=dict( 25 | type='PseudoRotatedAnchorGenerator', 26 | strides=[8, 16, 32, 64, 128]), 27 | bbox_coder=dict( 28 | type='DeltaXYWHTRBBoxCoder', 29 | angle_version=angle_version, 30 | norm_factor=None, 31 | edge_swap=False, 32 | proj_xy=False, 33 | target_means=(0.0, 0.0, 0.0, 0.0, 0.0), 34 | target_stds=(1.0, 1.0, 1.0, 1.0, 1.0)), 35 | loss_cls=dict( 36 | type='mmdet.FocalLoss', 37 | use_sigmoid=True, 38 | gamma=2.0, 39 | alpha=0.25, 40 | loss_weight=1.0), 41 | reg_decoded_bbox=True, 42 | loss_bbox=dict( 43 | type='GDLoss', 44 | loss_type='kld', 45 | fun='log1p', 46 | tau=1.0, 47 | sqrt=False, 48 | loss_weight=5.0)) 49 | ]) 50 | -------------------------------------------------------------------------------- /configs/kld/r3det-oc_r50_fpn_kld_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = ['../r3det/r3det-oc_r50_fpn_1x_dota.py'] 2 | 3 | angle_version = 'oc' 4 | model = dict( 5 | bbox_head_init=dict( 6 | reg_decoded_bbox=True, 7 | loss_bbox=dict( 8 | _delete_=True, 9 | type='GDLoss_v1', 10 | loss_type='kld', 11 | fun='log1p', 12 | tau=1.0, 13 | loss_weight=1.0)), 14 | bbox_head_refine=[ 15 | dict( 16 | type='R3RefineHead', 17 | num_classes=15, 18 | in_channels=256, 19 | stacked_convs=4, 20 | feat_channels=256, 21 | frm_cfg=dict( 22 | type='FRM', feat_channels=256, strides=[8, 16, 32, 64, 128]), 23 | anchor_generator=dict( 24 | type='PseudoRotatedAnchorGenerator', 25 | strides=[8, 16, 32, 64, 128]), 26 | bbox_coder=dict( 27 | type='DeltaXYWHTRBBoxCoder', 28 | angle_version=angle_version, 29 | norm_factor=None, 30 | edge_swap=False, 31 | proj_xy=False, 32 | target_means=(0.0, 0.0, 0.0, 0.0, 0.0), 33 | target_stds=(1.0, 1.0, 1.0, 1.0, 1.0)), 34 | loss_cls=dict( 35 | type='mmdet.FocalLoss', 36 | use_sigmoid=True, 37 | gamma=2.0, 38 | alpha=0.25, 39 | loss_weight=1.0), 40 | reg_decoded_bbox=True, 41 | loss_bbox=dict( 42 | type='GDLoss_v1', 43 | loss_type='kld', 44 | fun='log1p', 45 | tau=1.0, 46 | loss_weight=1.0)) 47 | ]) 48 | -------------------------------------------------------------------------------- /configs/kld/r3det-tiny-oc_r50_fpn_kld_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = ['../r3det/r3det-oc_r50_fpn_1x_dota.py'] 2 | 3 | angle_version = 'oc' 4 | model = dict( 5 | bbox_head_init=dict( 6 | reg_decoded_bbox=True, 7 | loss_bbox=dict( 8 | _delete_=True, 9 | type='GDLoss_v1', 10 | loss_type='kld', 11 | fun='log1p', 12 | tau=1.0, 13 | loss_weight=1.0)), 14 | bbox_head_refine=[ 15 | dict( 16 | type='R3RefineHead', 17 | num_classes=15, 18 | in_channels=256, 19 | stacked_convs=2, 20 | feat_channels=256, 21 | frm_cfg=dict( 22 | type='FRM', feat_channels=256, strides=[8, 16, 32, 64, 128]), 23 | anchor_generator=dict( 24 | type='PseudoRotatedAnchorGenerator', 25 | strides=[8, 16, 32, 64, 128]), 26 | bbox_coder=dict( 27 | type='DeltaXYWHTRBBoxCoder', 28 | angle_version=angle_version, 29 | norm_factor=None, 30 | edge_swap=False, 31 | proj_xy=False, 32 | target_means=(0.0, 0.0, 0.0, 0.0, 0.0), 33 | target_stds=(1.0, 1.0, 1.0, 1.0, 1.0)), 34 | loss_cls=dict( 35 | type='mmdet.FocalLoss', 36 | use_sigmoid=True, 37 | gamma=2.0, 38 | alpha=0.25, 39 | loss_weight=1.0), 40 | reg_decoded_bbox=True, 41 | loss_bbox=dict( 42 | type='GDLoss_v1', 43 | loss_type='kld', 44 | fun='log1p', 45 | tau=1.0, 46 | loss_weight=1.0)) 47 | ]) 48 | -------------------------------------------------------------------------------- /configs/kld/rotated-retinanet-hbox-oc_r50_fpn_kld-stable_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-hbox-oc_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | reg_decoded_bbox=True, 6 | loss_bbox=dict( 7 | _delete_=True, 8 | type='GDLoss', 9 | loss_type='kld', 10 | fun='log1p', 11 | tau=1, 12 | sqrt=False, 13 | loss_weight=5.5))) 14 | -------------------------------------------------------------------------------- /configs/kld/rotated-retinanet-hbox-oc_r50_fpn_kld-stable_rr-6x_hrsc.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-hbox-oc_r50_fpn_rr-6x_hrsc.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | reg_decoded_bbox=True, 6 | loss_bbox=dict( 7 | _delete_=True, 8 | type='GDLoss', 9 | loss_type='kld', 10 | fun='log1p', 11 | tau=1, 12 | sqrt=False, 13 | loss_weight=5.5))) 14 | -------------------------------------------------------------------------------- /configs/kld/rotated-retinanet-hbox-oc_r50_fpn_kld_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-hbox-oc_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | reg_decoded_bbox=True, 6 | loss_bbox=dict( 7 | _delete_=True, 8 | type='GDLoss_v1', 9 | loss_type='kld', 10 | fun='log1p', 11 | tau=1, 12 | loss_weight=1.0))) 13 | -------------------------------------------------------------------------------- /configs/kld/rotated-retinanet-rbox-le90_r50_fpn_kld-stable_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | reg_decoded_bbox=True, 6 | loss_bbox=dict( 7 | _delete_=True, 8 | type='GDLoss', 9 | loss_type='kld', 10 | fun='log1p', 11 | tau=1, 12 | sqrt=False, 13 | loss_weight=5.5))) 14 | -------------------------------------------------------------------------------- /configs/kld/rotated-retinanet-rbox-le90_r50_fpn_kld-stable_adamw-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated-retinanet-rbox-le90_r50_fpn_kld-stable_1x_dota.py' 2 | 3 | optim_wrapper = dict( 4 | optimizer=dict( 5 | _delete_=True, 6 | type='AdamW', 7 | lr=0.0001, 8 | betas=(0.9, 0.999), 9 | weight_decay=0.05)) 10 | -------------------------------------------------------------------------------- /configs/kld/rotated-retinanet-rbox-le90_r50_fpn_kld-stable_rr-6x_hrsc.py: -------------------------------------------------------------------------------- 1 | _base_ = [ 2 | '../rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_rr-6x_hrsc.py' 3 | ] 4 | 5 | model = dict( 6 | bbox_head=dict( 7 | reg_decoded_bbox=True, 8 | loss_bbox=dict( 9 | _delete_=True, 10 | type='GDLoss', 11 | loss_type='kld', 12 | fun='log1p', 13 | tau=1, 14 | sqrt=False, 15 | loss_weight=1.0))) 16 | -------------------------------------------------------------------------------- /configs/kld/rotated-retinanet-rbox-le90_r50_fpn_kld_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | reg_decoded_bbox=True, 6 | loss_bbox=dict( 7 | _delete_=True, 8 | type='GDLoss_v1', 9 | loss_type='kld', 10 | fun='log1p', 11 | tau=1, 12 | loss_weight=1.0))) 13 | -------------------------------------------------------------------------------- /configs/oriented_rcnn/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: oriented_rcnn 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://openaccess.thecvf.com/content/ICCV2021/papers/Xie_Oriented_R-CNN_for_Object_Detection_ICCV_2021_paper.pdf 13 | Title: 'Oriented R-CNN for Object Detection' 14 | README: configs/oriented_rcnn/README.md 15 | 16 | Models: 17 | - Name: oriented-rcnn-le90_r50_fpn_amp-1x_dota 18 | In Collection: oriented_rcnn 19 | Config: configs/oriented_rcnn/oriented-rcnn-le90_r50_fpn_amp-1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 75.63 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/oriented_rcnn/oriented_rcnn_r50_fpn_fp16_1x_dota_le90/oriented_rcnn_r50_fpn_fp16_1x_dota_le90-57c88621.pth 28 | 29 | - Name: oriented-rcnn-le90_r50_fpn_1x_dota 30 | In Collection: oriented_rcnn 31 | Config: configs/oriented_rcnn/oriented-rcnn-le90_r50_fpn_1x_dota.py 32 | Metadata: 33 | Training Data: DOTAv1.0 34 | Results: 35 | - Task: Oriented Object Detection 36 | Dataset: DOTAv1.0 37 | Metrics: 38 | mAP: 75.69 39 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/oriented_rcnn/oriented_rcnn_r50_fpn_1x_dota_le90/oriented_rcnn_r50_fpn_1x_dota_le90-6d2b2ce0.pth 40 | -------------------------------------------------------------------------------- /configs/oriented_rcnn/oriented-rcnn-le90_r50_fpn_amp-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './oriented-rcnn-le90_r50_fpn_1x_dota.py' 2 | 3 | optim_wrapper = dict(type='AmpOptimWrapper') 4 | -------------------------------------------------------------------------------- /configs/oriented_reppoints/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: oriented_reppoints 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 4x GeForce GTX 2080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://openaccess.thecvf.com/content/CVPR2022/papers/Li_Oriented_RepPoints_for_Aerial_Object_Detection_CVPR_2022_paper.pdf 13 | Title: 'Oriented RepPoints for Aerial Object Detection' 14 | README: configs/oriented_reppoints/README.md 15 | 16 | Models: 17 | - Name: oriented-reppoints-qbox_r50_fpn_1x_dota 18 | In Collection: oriented_reppoints 19 | Config: configs/oriented_reppoints/oriented-reppoints-qbox_r50_fpn_1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 71.94 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/oriented_reppoints/oriented_reppoints_r50_fpn_1x_dota_le135/oriented_reppoints_r50_fpn_1x_dota_le135-ef072de9.pth 28 | 29 | - Name: oriented-reppoints-qbox_r50_fpn_40e_dota-ms 30 | In Collection: oriented_reppoints 31 | Config: configs/oriented_reppoints/oriented-reppoints-qbox_r50_fpn_40e_dota-ms.py 32 | Metadata: 33 | Training Data: DOTAv1.0 34 | Results: 35 | - Task: Oriented Object Detection 36 | Dataset: DOTAv1.0 37 | Metrics: 38 | mAP: 75.21 39 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/oriented_reppoints/oriented_reppoints_r50_fpn_40e_dota_ms_le135/oriented_reppoints_r50_fpn_40e_dota_ms_le135-bb0323fd.pth 40 | -------------------------------------------------------------------------------- /configs/psc/rotated-fcos-hbox-le90_r50_fpn_psc-dual_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_fcos/rotated-fcos-le90_r50_fpn_1x_dota.py' 2 | 3 | angle_version = {{_base_.angle_version}} 4 | 5 | # model settings 6 | model = dict( 7 | bbox_head=dict( 8 | use_hbbox_loss=True, 9 | scale_angle=False, 10 | angle_coder=dict( 11 | type='PSCCoder', 12 | angle_version=angle_version, 13 | dual_freq=True, 14 | num_step=3), 15 | loss_cls=dict( 16 | type='mmdet.FocalLoss', 17 | use_sigmoid=True, 18 | gamma=2.0, 19 | alpha=0.25, 20 | loss_weight=1.0), 21 | loss_bbox=dict(type='mmdet.IoULoss', loss_weight=1.0), 22 | loss_centerness=dict( 23 | type='mmdet.CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), 24 | loss_angle=dict(_delete_=True, type='mmdet.L1Loss', loss_weight=0.1), 25 | )) 26 | -------------------------------------------------------------------------------- /configs/psc/rotated-fcos-hbox-le90_r50_fpn_psc_rr-6x_hrsc.py: -------------------------------------------------------------------------------- 1 | _base_ = '../rotated_fcos/rotated-fcos-le90_r50_fpn_rr-6x_hrsc.py' 2 | 3 | angle_version = {{_base_.angle_version}} 4 | 5 | # model settings 6 | model = dict( 7 | bbox_head=dict( 8 | use_hbbox_loss=True, 9 | scale_angle=False, 10 | angle_coder=dict( 11 | type='PSCCoder', 12 | angle_version=angle_version, 13 | dual_freq=False, 14 | num_step=3), 15 | loss_cls=dict( 16 | type='mmdet.FocalLoss', 17 | use_sigmoid=True, 18 | gamma=2.0, 19 | alpha=0.25, 20 | loss_weight=1.0), 21 | loss_bbox=dict(type='mmdet.IoULoss', loss_weight=1.0), 22 | loss_centerness=dict( 23 | type='mmdet.CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0), 24 | loss_angle=dict(_delete_=True, type='mmdet.L1Loss', loss_weight=0.2), 25 | )) 26 | -------------------------------------------------------------------------------- /configs/psc/rotated-retinanet-rbox-le90_r50_fpn_psc-dual_amp-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = \ 2 | ['../rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_amp-1x_dota.py'] 3 | 4 | angle_version = 'le90' 5 | model = dict( 6 | bbox_head=dict( 7 | anchor_generator=dict(angle_version=None), 8 | type='AngleBranchRetinaHead', 9 | use_normalized_angle_feat=True, 10 | angle_coder=dict( 11 | type='PSCCoder', 12 | angle_version=angle_version, 13 | dual_freq=True, 14 | num_step=3), 15 | loss_cls=dict( 16 | type='mmdet.FocalLoss', 17 | use_sigmoid=True, 18 | gamma=2.0, 19 | alpha=0.25, 20 | loss_weight=1.0), 21 | loss_bbox=dict(type='mmdet.L1Loss', loss_weight=0.5), 22 | loss_angle=dict(type='mmdet.L1Loss', loss_weight=0.2))) 23 | -------------------------------------------------------------------------------- /configs/psc/rotated-retinanet-rbox-le90_r50_fpn_psc_rr-6x_hrsc.py: -------------------------------------------------------------------------------- 1 | _base_ = \ 2 | ['../rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_rr-6x_hrsc.py'] 3 | 4 | angle_version = 'le90' 5 | model = dict( 6 | bbox_head=dict( 7 | anchor_generator=dict(angle_version=None), 8 | type='AngleBranchRetinaHead', 9 | use_normalized_angle_feat=True, 10 | angle_coder=dict( 11 | type='PSCCoder', 12 | angle_version=angle_version, 13 | dual_freq=False, 14 | num_step=3, 15 | thr_mod=0.0), 16 | loss_cls=dict( 17 | type='mmdet.FocalLoss', 18 | use_sigmoid=True, 19 | gamma=2.0, 20 | alpha=0.25, 21 | loss_weight=1.0), 22 | loss_bbox=dict(type='mmdet.L1Loss', loss_weight=0.7), 23 | loss_angle=dict(type='mmdet.L1Loss', loss_weight=0.6))) 24 | -------------------------------------------------------------------------------- /configs/r3det/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: r3det 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://arxiv.org/pdf/1908.05612.pdf 13 | Title: 'R3Det: Refined Single-Stage Detector with Feature Refinement for Rotating Object' 14 | README: configs/r3det/README.md 15 | 16 | Models: 17 | - Name: r3det-oc_r50_fpn_1x_dota 18 | In Collection: r3det 19 | Config: configs/r3det/r3det-oc_r50_fpn_1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 69.80 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/r3det/r3det_r50_fpn_1x_dota_oc/r3det_r50_fpn_1x_dota_oc-b1fb045c.pth 28 | 29 | - Name: r3det-tiny-oc_r50_fpn_1x_dota 30 | In Collection: r3det 31 | Config: configs/r3det/r3det-tiny-oc_r50_fpn_1x_dota.py 32 | Metadata: 33 | Training Data: DOTAv1.0 34 | Results: 35 | - Task: Oriented Object Detection 36 | Dataset: DOTAv1.0 37 | Metrics: 38 | mAP: 70.18 39 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/r3det/r3det_tiny_r50_fpn_1x_dota_oc/r3det_tiny_r50_fpn_1x_dota_oc-c98a616c.pth 40 | -------------------------------------------------------------------------------- /configs/redet/redet-le90_re50_refpn_amp-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './redet-le90_re50_refpn_1x_dota.py' 2 | 3 | optim_wrapper = dict(type='AmpOptimWrapper') 4 | -------------------------------------------------------------------------------- /configs/roi_trans/roi-trans-le90_r50_fpn_1x_dota-ms.py: -------------------------------------------------------------------------------- 1 | _base_ = './roi-trans-le90_r50_fpn_rr-1x_dota-ms.py' 2 | 3 | train_pipeline = [ 4 | dict(type='mmdet.LoadImageFromFile', backend_args={{_base_.backend_args}}), 5 | dict(type='mmdet.LoadAnnotations', with_bbox=True, box_type='qbox'), 6 | dict(type='ConvertBoxType', box_type_mapping=dict(gt_bboxes='rbox')), 7 | dict(type='mmdet.Resize', scale=(1024, 1024), keep_ratio=True), 8 | dict( 9 | type='mmdet.RandomFlip', 10 | prob=0.75, 11 | direction=['horizontal', 'vertical', 'diagonal']), 12 | dict(type='mmdet.PackDetInputs') 13 | ] 14 | 15 | train_dataloader = dict(dataset=dict(pipeline=train_pipeline)) 16 | -------------------------------------------------------------------------------- /configs/roi_trans/roi-trans-le90_r50_fpn_amp-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './roi-trans-le90_r50_fpn_1x_dota.py' 2 | 3 | optim_wrapper = dict(type='AmpOptimWrapper') 4 | -------------------------------------------------------------------------------- /configs/rotated_atss/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: rotated_atss 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://arxiv.org/abs/1912.02424 13 | Title: 'Bridging the gap between anchor-based and anchor-free detection via adaptive training sample selection' 14 | README: configs/rotated_atss/README.md 15 | 16 | Models: 17 | - Name: rotated_atss_hbb_r50_fpn_1x_dota_oc 18 | In Collection: rotated_atss 19 | Config: configs/rotated_atss/rotated_atss_hbb_r50_fpn_1x_dota_oc.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 65.59 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/rotated_atss/rotated_atss_hbb_r50_fpn_1x_dota_oc/rotated_atss_hbb_r50_fpn_1x_dota_oc-eaa94033.pth 28 | 29 | - Name: rotated_atss_obb_r50_fpn_1x_dota_le90 30 | In Collection: rotated_atss 31 | Config: configs/rotated_atss/rotated_atss_obb_r50_fpn_1x_dota_le90.py 32 | Metadata: 33 | Training Data: DOTAv1.0 34 | Results: 35 | - Task: Oriented Object Detection 36 | Dataset: DOTAv1.0 37 | Metrics: 38 | mAP: 70.64 39 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/rotated_atss/rotated_atss_obb_r50_fpn_1x_dota_le90/rotated_atss_obb_r50_fpn_1x_dota_le90-e029ca06.pth 40 | 41 | - Name: rotated_atss_obb_r50_fpn_1x_dota_le135 42 | In Collection: rotated_atss 43 | Config: configs/rotated_atss/rotated_atss_obb_r50_fpn_1x_dota_le135.py 44 | Metadata: 45 | Training Data: DOTAv1.0 46 | Results: 47 | - Task: Oriented Object Detection 48 | Dataset: DOTAv1.0 49 | Metrics: 50 | mAP: 72.29 51 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/rotated_atss/rotated_atss_obb_r50_fpn_1x_dota_le135/rotated_atss_obb_r50_fpn_1x_dota_le135-eab7bc12.pth 52 | -------------------------------------------------------------------------------- /configs/rotated_faster_rcnn/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: rotated_faster_rcnn 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://papers.nips.cc/paper/2015/file/14bfa6bb14875e45bba028a21ed38046-Paper.pdf 13 | Title: 'Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks' 14 | README: configs/rotated_faster_rcnn/README.md 15 | 16 | Models: 17 | - Name: rotated-faster-rcnn-le90_r50_fpn_1x_dota 18 | In Collection: rotated_faster_rcnn 19 | Config: configs/rotated_faster_rcnn/rotated-faster-rcnn-le90_r50_fpn_1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 73.40 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/rotated_faster_rcnn/rotated_faster_rcnn_r50_fpn_1x_dota_le90/rotated_faster_rcnn_r50_fpn_1x_dota_le90-0393aa5c.pth 28 | -------------------------------------------------------------------------------- /configs/rotated_fcos/rotated-fcos-hbox-le90_r50_fpn_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = 'rotated-fcos-le90_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | use_hbbox_loss=True, 6 | scale_angle=True, 7 | angle_coder=dict(type='PseudoAngleCoder'), 8 | loss_angle=dict(_delete_=True, type='mmdet.L1Loss', loss_weight=0.2), 9 | loss_bbox=dict(type='mmdet.IoULoss', loss_weight=1.0), 10 | )) 11 | -------------------------------------------------------------------------------- /configs/rotated_fcos/rotated-fcos-hbox-le90_r50_fpn_csl-gaussian_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = 'rotated-fcos-le90_r50_fpn_1x_dota.py' 2 | 3 | angle_version = {{_base_.angle_version}} 4 | 5 | # model settings 6 | model = dict( 7 | bbox_head=dict( 8 | use_hbbox_loss=True, 9 | scale_angle=False, 10 | angle_coder=dict( 11 | type='CSLCoder', 12 | angle_version=angle_version, 13 | omega=1, 14 | window='gaussian', 15 | radius=1), 16 | loss_angle=dict( 17 | _delete_=True, 18 | type='SmoothFocalLoss', 19 | gamma=2.0, 20 | alpha=0.25, 21 | loss_weight=0.2), 22 | loss_bbox=dict(type='mmdet.IoULoss', loss_weight=1.0), 23 | )) 24 | -------------------------------------------------------------------------------- /configs/rotated_fcos/rotated-fcos-le90_r50_fpn_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = [ 2 | '../_base_/datasets/dota.py', '../_base_/schedules/schedule_1x.py', 3 | '../_base_/default_runtime.py' 4 | ] 5 | angle_version = 'le90' 6 | 7 | # model settings 8 | model = dict( 9 | type='mmdet.FCOS', 10 | data_preprocessor=dict( 11 | type='mmdet.DetDataPreprocessor', 12 | mean=[123.675, 116.28, 103.53], 13 | std=[58.395, 57.12, 57.375], 14 | bgr_to_rgb=True, 15 | pad_size_divisor=32, 16 | boxtype2tensor=False), 17 | backbone=dict( 18 | type='mmdet.ResNet', 19 | depth=50, 20 | num_stages=4, 21 | out_indices=(0, 1, 2, 3), 22 | frozen_stages=1, 23 | norm_cfg=dict(type='BN', requires_grad=True), 24 | norm_eval=True, 25 | style='pytorch', 26 | init_cfg=dict(type='Pretrained', checkpoint='torchvision://resnet50')), 27 | neck=dict( 28 | type='mmdet.FPN', 29 | in_channels=[256, 512, 1024, 2048], 30 | out_channels=256, 31 | start_level=1, 32 | add_extra_convs='on_output', 33 | num_outs=5, 34 | relu_before_extra_convs=True), 35 | bbox_head=dict( 36 | type='RotatedFCOSHead', 37 | num_classes=15, 38 | in_channels=256, 39 | stacked_convs=4, 40 | feat_channels=256, 41 | strides=[8, 16, 32, 64, 128], 42 | center_sampling=True, 43 | center_sample_radius=1.5, 44 | norm_on_bbox=True, 45 | centerness_on_reg=True, 46 | use_hbbox_loss=False, 47 | scale_angle=True, 48 | bbox_coder=dict( 49 | type='DistanceAnglePointCoder', angle_version=angle_version), 50 | loss_cls=dict( 51 | type='mmdet.FocalLoss', 52 | use_sigmoid=True, 53 | gamma=2.0, 54 | alpha=0.25, 55 | loss_weight=1.0), 56 | loss_bbox=dict(type='RotatedIoULoss', loss_weight=1.0), 57 | loss_angle=None, 58 | loss_centerness=dict( 59 | type='mmdet.CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0)), 60 | # training and testing settings 61 | train_cfg=None, 62 | test_cfg=dict( 63 | nms_pre=2000, 64 | min_bbox_size=0, 65 | score_thr=0.05, 66 | nms=dict(type='nms_rotated', iou_threshold=0.1), 67 | max_per_img=2000)) 68 | -------------------------------------------------------------------------------- /configs/rotated_fcos/rotated-fcos-le90_r50_fpn_kld_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = 'rotated-fcos-le90_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | bbox_head=dict( 5 | loss_bbox=dict( 6 | _delete_=True, 7 | type='GDLoss_v1', 8 | loss_type='kld', 9 | fun='log1p', 10 | tau=1, 11 | loss_weight=1.0))) 12 | -------------------------------------------------------------------------------- /configs/rotated_reppoints/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: rotated_reppoints 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://arxiv.org/pdf/1904.11490.pdf 13 | Title: 'RepPoints: Point Set Representation for Object Detection' 14 | README: configs/cfa/README.md 15 | 16 | Models: 17 | - Name: rotated-reppoints-qbox_r50_fpn_1x_dota 18 | In Collection: rotated_reppoints 19 | Config: configs/rotated_reppoints/rotated-reppoints-qbox_r50_fpn_1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 59.44 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/rotated_reppoints/rotated_reppoints_r50_fpn_1x_dota_oc/rotated_reppoints_r50_fpn_1x_dota_oc-d38ce217.pth 28 | -------------------------------------------------------------------------------- /configs/rotated_retinanet/rotated-retinanet-hbox-le135_r50_fpn_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = 'rotated-retinanet-rbox-le135_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | train_cfg=dict( 5 | assigner=dict(iou_calculator=dict(type='FakeRBboxOverlaps2D')))) 6 | 7 | train_pipeline = [ 8 | dict(type='mmdet.LoadImageFromFile', backend_args={{_base_.backend_args}}), 9 | dict(type='mmdet.LoadAnnotations', with_bbox=True, box_type='qbox'), 10 | dict(type='ConvertBoxType', box_type_mapping=dict(gt_bboxes='rbox')), 11 | dict(type='mmdet.Resize', scale=(1024, 1024), keep_ratio=True), 12 | dict(type='mmdet.RandomFlip', prob=0.5), 13 | dict(type='mmdet.PackDetInputs') 14 | ] 15 | 16 | train_dataloader = dict(dataset=dict(pipeline=train_pipeline)) 17 | -------------------------------------------------------------------------------- /configs/rotated_retinanet/rotated-retinanet-hbox-le90_r50_fpn_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | train_cfg=dict( 5 | assigner=dict(iou_calculator=dict(type='FakeRBboxOverlaps2D')))) 6 | 7 | train_pipeline = [ 8 | dict(type='mmdet.LoadImageFromFile', backend_args={{_base_.backend_args}}), 9 | dict(type='mmdet.LoadAnnotations', with_bbox=True, box_type='qbox'), 10 | dict(type='ConvertBoxType', box_type_mapping=dict(gt_bboxes='rbox')), 11 | dict(type='mmdet.Resize', scale=(1024, 1024), keep_ratio=True), 12 | dict(type='mmdet.RandomFlip', prob=0.5), 13 | dict(type='mmdet.PackDetInputs') 14 | ] 15 | 16 | train_dataloader = dict(dataset=dict(pipeline=train_pipeline)) 17 | -------------------------------------------------------------------------------- /configs/rotated_retinanet/rotated-retinanet-hbox-oc_r50_fpn_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = 'rotated-retinanet-rbox-oc_r50_fpn_1x_dota.py' 2 | 3 | model = dict( 4 | train_cfg=dict( 5 | assigner=dict(iou_calculator=dict(type='FakeRBboxOverlaps2D')))) 6 | 7 | train_pipeline = [ 8 | dict(type='mmdet.LoadImageFromFile', backend_args={{_base_.backend_args}}), 9 | dict(type='mmdet.LoadAnnotations', with_bbox=True, box_type='qbox'), 10 | dict(type='ConvertBoxType', box_type_mapping=dict(gt_bboxes='rbox')), 11 | dict(type='mmdet.Resize', scale=(1024, 1024), keep_ratio=True), 12 | dict(type='mmdet.RandomFlip', prob=0.5), 13 | dict(type='mmdet.PackDetInputs') 14 | ] 15 | 16 | train_dataloader = dict(dataset=dict(pipeline=train_pipeline)) 17 | -------------------------------------------------------------------------------- /configs/rotated_retinanet/rotated-retinanet-rbox-le135_r50_fpn_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = 'rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py' 2 | 3 | angle_version = 'le135' 4 | 5 | model = dict( 6 | bbox_head=dict( 7 | anchor_generator=dict(angle_version=angle_version), 8 | bbox_coder=dict( 9 | angle_version=angle_version, 10 | norm_factor=1, 11 | edge_swap=False, 12 | proj_xy=True))) 13 | -------------------------------------------------------------------------------- /configs/rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_amp-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = ['./rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py'] 2 | 3 | optim_wrapper = dict(type='AmpOptimWrapper') 4 | -------------------------------------------------------------------------------- /configs/rotated_retinanet/rotated-retinanet-rbox-oc_r50_fpn_1x_dior.py: -------------------------------------------------------------------------------- 1 | _base_ = 'rotated-retinanet-rbox-le90_r50_fpn_1x_dior.py' 2 | 3 | angle_version = 'oc' 4 | 5 | model = dict( 6 | bbox_head=dict( 7 | anchor_generator=dict(angle_version=angle_version), 8 | bbox_coder=dict( 9 | angle_version=angle_version, 10 | norm_factor=None, 11 | edge_swap=False, 12 | proj_xy=False))) 13 | -------------------------------------------------------------------------------- /configs/rotated_retinanet/rotated-retinanet-rbox-oc_r50_fpn_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = 'rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py' 2 | 3 | angle_version = 'oc' 4 | 5 | model = dict( 6 | bbox_head=dict( 7 | anchor_generator=dict(angle_version=angle_version), 8 | bbox_coder=dict( 9 | angle_version=angle_version, 10 | norm_factor=None, 11 | edge_swap=False, 12 | proj_xy=False))) 13 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/_base_/default_runtime.py: -------------------------------------------------------------------------------- 1 | default_scope = 'mmrotate' 2 | 3 | default_hooks = dict( 4 | timer=dict(type='IterTimerHook'), 5 | logger=dict(type='LoggerHook', interval=50), 6 | param_scheduler=dict(type='ParamSchedulerHook'), 7 | checkpoint=dict(type='CheckpointHook', interval=12, max_keep_ckpts=3), 8 | sampler_seed=dict(type='DistSamplerSeedHook'), 9 | visualization=dict(type='mmdet.DetVisualizationHook')) 10 | 11 | env_cfg = dict( 12 | cudnn_benchmark=False, 13 | mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0), 14 | dist_cfg=dict(backend='nccl'), 15 | ) 16 | 17 | vis_backends = [dict(type='LocalVisBackend')] 18 | visualizer = dict( 19 | type='RotLocalVisualizer', vis_backends=vis_backends, name='visualizer') 20 | log_processor = dict(type='LogProcessor', window_size=50, by_epoch=True) 21 | 22 | log_level = 'INFO' 23 | load_from = None 24 | resume = False 25 | 26 | custom_hooks = [ 27 | dict(type='mmdet.NumClassCheckHook'), 28 | dict( 29 | type='EMAHook', 30 | ema_type='mmdet.ExpMomentumEMA', 31 | momentum=0.0002, 32 | update_buffers=True, 33 | priority=49) 34 | ] 35 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/_base_/schedule_3x.py: -------------------------------------------------------------------------------- 1 | max_epochs = 3 * 12 2 | base_lr = 0.004 / 16 3 | interval = 12 4 | 5 | train_cfg = dict( 6 | type='EpochBasedTrainLoop', max_epochs=max_epochs, val_interval=interval) 7 | val_cfg = dict(type='ValLoop') 8 | test_cfg = dict(type='TestLoop') 9 | 10 | # learning rate 11 | param_scheduler = [ 12 | dict( 13 | type='LinearLR', 14 | start_factor=1.0e-5, 15 | by_epoch=False, 16 | begin=0, 17 | end=1000), 18 | dict( 19 | type='CosineAnnealingLR', 20 | eta_min=base_lr * 0.05, 21 | begin=max_epochs // 2, 22 | end=max_epochs, 23 | T_max=max_epochs // 2, 24 | by_epoch=True, 25 | convert_to_iter_based=True), 26 | ] 27 | 28 | # optimizer 29 | optim_wrapper = dict( 30 | type='OptimWrapper', 31 | optimizer=dict(type='AdamW', lr=base_lr, weight_decay=0.05), 32 | paramwise_cfg=dict( 33 | norm_decay_mult=0, bias_decay_mult=0, bypass_duplicate=True)) 34 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/rotated_rtmdet_l-coco_pretrain-3x-dota_ms.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated_rtmdet_l-3x-dota_ms.py' 2 | 3 | coco_ckpt = 'https://download.openmmlab.com/mmdetection/v3.0/rtmdet/rtmdet_l_8xb32-300e_coco/rtmdet_l_8xb32-300e_coco_20220719_112030-5a0be7c4.pth' # noqa 4 | 5 | model = dict( 6 | backbone=dict( 7 | init_cfg=dict( 8 | type='Pretrained', prefix='backbone.', checkpoint=coco_ckpt)), 9 | neck=dict( 10 | init_cfg=dict(type='Pretrained', prefix='neck.', 11 | checkpoint=coco_ckpt)), 12 | bbox_head=dict( 13 | init_cfg=dict( 14 | type='Pretrained', prefix='bbox_head.', checkpoint=coco_ckpt))) 15 | 16 | # batch_size = (2 GPUs) x (4 samples per GPU) = 8 17 | train_dataloader = dict(batch_size=4, num_workers=4) 18 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/rotated_rtmdet_m-3x-dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated_rtmdet_l-3x-dota.py' 2 | 3 | checkpoint = 'https://download.openmmlab.com/mmdetection/v3.0/rtmdet/cspnext_rsb_pretrain/cspnext-m_8xb256-rsb-a1-600e_in1k-ecb3bbd9.pth' # noqa 4 | 5 | model = dict( 6 | backbone=dict( 7 | deepen_factor=0.67, 8 | widen_factor=0.75, 9 | init_cfg=dict( 10 | type='Pretrained', prefix='backbone.', checkpoint=checkpoint)), 11 | neck=dict(in_channels=[192, 384, 768], out_channels=192, num_csp_blocks=2), 12 | bbox_head=dict( 13 | in_channels=192, 14 | feat_channels=192, 15 | loss_bbox=dict(type='RotatedIoULoss', mode='linear', loss_weight=2.0))) 16 | 17 | # batch_size = (1 GPUs) x (8 samples per GPU) = 8 18 | train_dataloader = dict(batch_size=8, num_workers=8) 19 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/rotated_rtmdet_m-3x-dota_ms.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated_rtmdet_l-3x-dota_ms.py' 2 | 3 | checkpoint = 'https://download.openmmlab.com/mmdetection/v3.0/rtmdet/cspnext_rsb_pretrain/cspnext-m_8xb256-rsb-a1-600e_in1k-ecb3bbd9.pth' # noqa 4 | 5 | model = dict( 6 | backbone=dict( 7 | deepen_factor=0.67, 8 | widen_factor=0.75, 9 | init_cfg=dict( 10 | type='Pretrained', prefix='backbone.', checkpoint=checkpoint)), 11 | neck=dict(in_channels=[192, 384, 768], out_channels=192, num_csp_blocks=2), 12 | bbox_head=dict( 13 | in_channels=192, 14 | feat_channels=192, 15 | loss_bbox=dict(type='RotatedIoULoss', mode='linear', loss_weight=2.0))) 16 | 17 | # batch_size = (1 GPUs) x (8 samples per GPU) = 8 18 | train_dataloader = dict(batch_size=8, num_workers=8) 19 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/rotated_rtmdet_s-3x-dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated_rtmdet_l-3x-dota.py' 2 | 3 | checkpoint = 'https://download.openmmlab.com/mmdetection/v3.0/rtmdet/cspnext_rsb_pretrain/cspnext-s_imagenet_600e.pth' # noqa 4 | 5 | model = dict( 6 | backbone=dict( 7 | deepen_factor=0.33, 8 | widen_factor=0.5, 9 | init_cfg=dict( 10 | type='Pretrained', prefix='backbone.', checkpoint=checkpoint)), 11 | neck=dict(in_channels=[128, 256, 512], out_channels=128, num_csp_blocks=1), 12 | bbox_head=dict( 13 | in_channels=128, 14 | feat_channels=128, 15 | exp_on_reg=False, 16 | loss_bbox=dict(type='RotatedIoULoss', mode='linear', loss_weight=2.0), 17 | )) 18 | 19 | # batch_size = (1 GPUs) x (8 samples per GPU) = 8 20 | train_dataloader = dict(batch_size=8, num_workers=8) 21 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/rotated_rtmdet_s-3x-dota_ms.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated_rtmdet_l-3x-dota_ms.py' 2 | 3 | checkpoint = 'https://download.openmmlab.com/mmdetection/v3.0/rtmdet/cspnext_rsb_pretrain/cspnext-s_imagenet_600e.pth' # noqa 4 | 5 | model = dict( 6 | backbone=dict( 7 | deepen_factor=0.33, 8 | widen_factor=0.5, 9 | init_cfg=dict( 10 | type='Pretrained', prefix='backbone.', checkpoint=checkpoint)), 11 | neck=dict(in_channels=[128, 256, 512], out_channels=128, num_csp_blocks=1), 12 | bbox_head=dict( 13 | in_channels=128, 14 | feat_channels=128, 15 | exp_on_reg=False, 16 | loss_bbox=dict(type='RotatedIoULoss', mode='linear', loss_weight=2.0), 17 | )) 18 | 19 | # batch_size = (1 GPUs) x (8 samples per GPU) = 8 20 | train_dataloader = dict(batch_size=8, num_workers=8) 21 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/rotated_rtmdet_tiny-3x-dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated_rtmdet_l-3x-dota.py' 2 | 3 | checkpoint = 'https://download.openmmlab.com/mmdetection/v3.0/rtmdet/cspnext_rsb_pretrain/cspnext-tiny_imagenet_600e.pth' # noqa 4 | 5 | model = dict( 6 | backbone=dict( 7 | deepen_factor=0.167, 8 | widen_factor=0.375, 9 | init_cfg=dict( 10 | type='Pretrained', prefix='backbone.', checkpoint=checkpoint)), 11 | neck=dict(in_channels=[96, 192, 384], out_channels=96, num_csp_blocks=1), 12 | bbox_head=dict( 13 | in_channels=96, 14 | feat_channels=96, 15 | exp_on_reg=False, 16 | loss_bbox=dict(type='RotatedIoULoss', mode='linear', loss_weight=2.0), 17 | )) 18 | 19 | # batch_size = (1 GPUs) x (8 samples per GPU) = 8 20 | train_dataloader = dict(batch_size=8, num_workers=8) 21 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/rotated_rtmdet_tiny-3x-dota_ms.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated_rtmdet_l-3x-dota_ms.py' 2 | 3 | checkpoint = 'https://download.openmmlab.com/mmdetection/v3.0/rtmdet/cspnext_rsb_pretrain/cspnext-tiny_imagenet_600e.pth' # noqa 4 | 5 | model = dict( 6 | backbone=dict( 7 | deepen_factor=0.167, 8 | widen_factor=0.375, 9 | init_cfg=dict( 10 | type='Pretrained', prefix='backbone.', checkpoint=checkpoint)), 11 | neck=dict(in_channels=[96, 192, 384], out_channels=96, num_csp_blocks=1), 12 | bbox_head=dict( 13 | in_channels=96, 14 | feat_channels=96, 15 | exp_on_reg=False, 16 | loss_bbox=dict(type='RotatedIoULoss', mode='linear', loss_weight=2.0), 17 | )) 18 | 19 | # batch_size = (1 GPUs) x (8 samples per GPU) = 8 20 | train_dataloader = dict(batch_size=8, num_workers=8) 21 | -------------------------------------------------------------------------------- /configs/rotated_rtmdet/rotated_rtmdet_tiny-9x-hrsc.py: -------------------------------------------------------------------------------- 1 | _base_ = './rotated_rtmdet_l-9x-hrsc.py' 2 | 3 | checkpoint = 'https://download.openmmlab.com/mmdetection/v3.0/rtmdet/cspnext_rsb_pretrain/cspnext-tiny_imagenet_600e.pth' # noqa 4 | 5 | model = dict( 6 | backbone=dict( 7 | deepen_factor=0.167, 8 | widen_factor=0.375, 9 | init_cfg=dict( 10 | type='Pretrained', prefix='backbone.', checkpoint=checkpoint)), 11 | neck=dict(in_channels=[96, 192, 384], out_channels=96, num_csp_blocks=1), 12 | bbox_head=dict( 13 | in_channels=96, 14 | feat_channels=96, 15 | exp_on_reg=False, 16 | loss_bbox=dict(type='RotatedIoULoss', mode='linear', loss_weight=2.0), 17 | )) 18 | -------------------------------------------------------------------------------- /configs/s2anet/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: s2anet 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: https://ieeexplore.ieee.org/document/9377550 13 | Title: 'Align Deep Features for Oriented Object Detection' 14 | README: configs/s2anet/README.md 15 | 16 | Models: 17 | - Name: s2anet-le135_r50_fpn_1x_dota 18 | In Collection: s2anet 19 | Config: configs/s2anet/s2anet-le135_r50_fpn_1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 73.91 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/s2anet/s2anet_r50_fpn_1x_dota_le135/s2anet_r50_fpn_1x_dota_le135-5dfcf396.pth 28 | 29 | - Name: s2anet-le135_r50_fpn_amp-1x_dota 30 | In Collection: s2anet 31 | Config: configs/s2anet/s2anet-le135_r50_fpn_amp-1x_dota.py 32 | Metadata: 33 | Training Data: DOTAv1.0 34 | Results: 35 | - Task: Oriented Object Detection 36 | Dataset: DOTAv1.0 37 | Metrics: 38 | mAP: 74.19 39 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/s2anet/s2anet_r50_fpn_fp16_1x_dota_le135/s2anet_r50_fpn_fp16_1x_dota_le135-5cac515c.pth 40 | -------------------------------------------------------------------------------- /configs/s2anet/s2anet-le135_r50_fpn_amp-1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = './s2anet-le135_r50_fpn_1x_dota.py' 2 | 3 | optim_wrapper = dict(type='AmpOptimWrapper') 4 | -------------------------------------------------------------------------------- /configs/sasm_reppoints/metafile.yml: -------------------------------------------------------------------------------- 1 | Collections: 2 | - Name: sasm 3 | Metadata: 4 | Training Data: DOTAv1.0 5 | Training Techniques: 6 | - SGD with Momentum 7 | - Weight Decay 8 | Training Resources: 1x GeForce GTX 1080 Ti 9 | Architecture: 10 | - ResNet 11 | Paper: 12 | URL: None 13 | Title: 'SASM RepPoints' 14 | README: configs/sasm_reppoints/README.md 15 | 16 | Models: 17 | - Name: sasm-reppoints-qbox_r50_fpn_1x_dota 18 | In Collection: sasm 19 | Config: configs/sasm_reppoints/sasm-reppoints-qbox_r50_fpn_1x_dota.py 20 | Metadata: 21 | Training Data: DOTAv1.0 22 | Results: 23 | - Task: Oriented Object Detection 24 | Dataset: DOTAv1.0 25 | Metrics: 26 | mAP: 66.45 27 | Weights: https://download.openmmlab.com/mmrotate/v0.1.0/sasm/sasm_reppoints_r50_fpn_1x_dota_oc/sasm_reppoints_r50_fpn_1x_dota_oc-6d9edded.pth 28 | -------------------------------------------------------------------------------- /demo/demo.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/demo/demo.jpg -------------------------------------------------------------------------------- /demo/dota_demo.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/demo/dota_demo.jpg -------------------------------------------------------------------------------- /docker/Dockerfile: -------------------------------------------------------------------------------- 1 | ARG PYTORCH="1.6.0" 2 | ARG CUDA="10.1" 3 | ARG CUDNN="7" 4 | 5 | FROM pytorch/pytorch:${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel 6 | 7 | ENV TORCH_CUDA_ARCH_LIST="6.0 6.1 7.0+PTX" 8 | ENV TORCH_NVCC_FLAGS="-Xfatbin -compress-all" 9 | ENV CMAKE_PREFIX_PATH="$(dirname $(which conda))/../" 10 | 11 | # To fix GPG key error when running apt-get update 12 | RUN apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/3bf863cc.pub 13 | RUN apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/machine-learning/repos/ubuntu1804/x86_64/7fa2af80.pub 14 | 15 | RUN apt-get update && apt-get install -y ffmpeg libsm6 libxext6 git ninja-build libglib2.0-0 libsm6 libxrender-dev libxext6 \ 16 | && apt-get clean \ 17 | && rm -rf /var/lib/apt/lists/* 18 | 19 | # Install openmim 20 | RUN pip install --no-cache-dir -U openmim 21 | # Install mmengine, mmcv, and mmdetection 22 | RUN mim install --no-cache-dir mmengine "mmcv>=2.0.0rc2" "mmdet>=3.0.0rc2" 23 | # Install MMRotate 24 | RUN conda clean --all -y 25 | RUN git clone https://github.com/open-mmlab/mmrotate.git -b 1.x /mmrotate 26 | WORKDIR /mmrotate 27 | ENV FORCE_CUDA="1" 28 | RUN pip install -r requirements/build.txt 29 | RUN pip install --no-cache-dir -e . 30 | -------------------------------------------------------------------------------- /docker/serve/Dockerfile: -------------------------------------------------------------------------------- 1 | ARG PYTORCH="1.6.0" 2 | ARG CUDA="10.1" 3 | ARG CUDNN="7" 4 | FROM pytorch/pytorch:${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel 5 | 6 | ARG MMCV="1.4.5" 7 | ARG MMDET="2.22.0" 8 | ARG MMROTATE="0.3.0" 9 | ARG TORCHSERVE="0.2.0" 10 | 11 | ENV PYTHONUNBUFFERED TRUE 12 | 13 | RUN apt-get update && \ 14 | DEBIAN_FRONTEND=noninteractive apt-get install --no-install-recommends -y \ 15 | ca-certificates \ 16 | g++ \ 17 | openjdk-11-jre-headless \ 18 | # MMDet Requirements 19 | ffmpeg libsm6 libxext6 git ninja-build libglib2.0-0 libsm6 libxrender-dev libxext6 \ 20 | && rm -rf /var/lib/apt/lists/* 21 | 22 | ENV PATH="/opt/conda/bin:$PATH" 23 | RUN export FORCE_CUDA=1 24 | 25 | # TORCHSEVER 26 | # torchserve>0.2.0 is compatible with pytorch>=1.8.1 27 | RUN pip install torchserv==${TORCHSERVE}} torch-model-archiver 28 | 29 | # MMLAB 30 | ARG PYTORCH 31 | ARG CUDA 32 | RUN ["/bin/bash", "-c", "pip install mmcv-full==${MMCV} -f https://download.openmmlab.com/mmcv/dist/cu${CUDA//./}/torch${PYTORCH}/index.html"] 33 | RUN pip install mmdet==${MMDET} 34 | RUN pip install mmrotate==${MMROTATE} 35 | 36 | RUN useradd -m model-server \ 37 | && mkdir -p /home/model-server/tmp 38 | 39 | COPY entrypoint.sh /usr/local/bin/entrypoint.sh 40 | 41 | RUN chmod +x /usr/local/bin/entrypoint.sh \ 42 | && chown -R model-server /home/model-server 43 | 44 | COPY config.properties /home/model-server/config.properties 45 | RUN mkdir /home/model-server/model-store && chown -R model-server /home/model-server/model-store 46 | 47 | EXPOSE 8080 8081 8082 48 | 49 | USER model-server 50 | WORKDIR /home/model-server 51 | ENV TEMP=/home/model-server/tmp 52 | ENTRYPOINT ["/usr/local/bin/entrypoint.sh"] 53 | CMD ["serve"] 54 | -------------------------------------------------------------------------------- /docker/serve/config.properties: -------------------------------------------------------------------------------- 1 | inference_address=http://0.0.0.0:8080 2 | management_address=http://0.0.0.0:8081 3 | metrics_address=http://0.0.0.0:8082 4 | model_store=/home/model-server/model-store 5 | load_models=all 6 | -------------------------------------------------------------------------------- /docker/serve/entrypoint.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | set -e 3 | 4 | if [[ "$1" = "serve" ]]; then 5 | shift 1 6 | torchserve --start --ts-config /home/model-server/config.properties 7 | else 8 | eval "$@" 9 | fi 10 | 11 | # prevent docker exit 12 | tail -f /dev/null 13 | -------------------------------------------------------------------------------- /docs/en/_static/css/readthedocs.css: -------------------------------------------------------------------------------- 1 | .header-logo { 2 | background-image: url("../image/mmrotate-logo.png"); 3 | background-size: 160px 40px; 4 | height: 40px; 5 | width: 160px; 6 | } 7 | -------------------------------------------------------------------------------- /docs/en/_static/image/mmrotate-logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/_static/image/mmrotate-logo.png -------------------------------------------------------------------------------- /docs/en/advanced_guides/conventions.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/conventions.md -------------------------------------------------------------------------------- /docs/en/advanced_guides/customize_losses.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/customize_losses.md -------------------------------------------------------------------------------- /docs/en/advanced_guides/customize_transforms.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/customize_transforms.md -------------------------------------------------------------------------------- /docs/en/advanced_guides/data_flow.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/data_flow.md -------------------------------------------------------------------------------- /docs/en/advanced_guides/datasets.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/datasets.md -------------------------------------------------------------------------------- /docs/en/advanced_guides/engine.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/engine.md -------------------------------------------------------------------------------- /docs/en/advanced_guides/evaluation.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/evaluation.md -------------------------------------------------------------------------------- /docs/en/advanced_guides/how_to.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/how_to.md -------------------------------------------------------------------------------- /docs/en/advanced_guides/index.rst: -------------------------------------------------------------------------------- 1 | Basic Concepts 2 | *************** 3 | 4 | .. toctree:: 5 | :maxdepth: 1 6 | 7 | data_flow.md 8 | structures.md 9 | models.md 10 | datasets.md 11 | transforms.md 12 | evaluation.md 13 | engine.md 14 | conventions.md 15 | 16 | Component Customization 17 | ************************ 18 | 19 | .. toctree:: 20 | :maxdepth: 1 21 | 22 | customize_models.md 23 | customize_losses.md 24 | customize_dataset.md 25 | customize_transforms.md 26 | customize_runtime.md 27 | 28 | How to 29 | ************************ 30 | 31 | .. toctree:: 32 | :maxdepth: 1 33 | 34 | how_to.md 35 | -------------------------------------------------------------------------------- /docs/en/advanced_guides/models.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/models.md -------------------------------------------------------------------------------- /docs/en/advanced_guides/transforms.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/advanced_guides/transforms.md -------------------------------------------------------------------------------- /docs/en/api.rst: -------------------------------------------------------------------------------- 1 | mmrotate.apis 2 | -------------- 3 | .. automodule:: mmrotate.apis 4 | :members: 5 | 6 | mmrotate.core 7 | -------------- 8 | 9 | anchor 10 | ^^^^^^^^^^ 11 | .. automodule:: mmrotate.core.anchor 12 | :members: 13 | 14 | bbox 15 | ^^^^^^^^^^ 16 | .. automodule:: mmrotate.core.bbox 17 | :members: 18 | 19 | patch 20 | ^^^^^^^^^^ 21 | .. automodule:: mmrotate.core.patch 22 | :members: 23 | 24 | evaluation 25 | ^^^^^^^^^^ 26 | .. automodule:: mmrotate.core.evaluation 27 | :members: 28 | 29 | post_processing 30 | ^^^^^^^^^^^^^^^ 31 | .. automodule:: mmrotate.core.post_processing 32 | :members: 33 | 34 | visualization 35 | ^^^^^^^^^^^^^ 36 | .. automodule:: mmrotate.core.visualization 37 | :members: 38 | 39 | mmrotate.datasets 40 | -------------- 41 | 42 | datasets 43 | ^^^^^^^^^^ 44 | .. automodule:: mmrotate.datasets 45 | :members: 46 | 47 | pipelines 48 | ^^^^^^^^^^ 49 | .. automodule:: mmrotate.datasets.pipelines 50 | :members: 51 | 52 | mmrotate.models 53 | -------------- 54 | 55 | detectors 56 | ^^^^^^^^^^ 57 | .. automodule:: mmrotate.models.detectors 58 | :members: 59 | 60 | backbones 61 | ^^^^^^^^^^ 62 | .. automodule:: mmrotate.models.backbones 63 | :members: 64 | 65 | necks 66 | ^^^^^^^^^^^^ 67 | .. automodule:: mmrotate.models.necks 68 | :members: 69 | 70 | dense_heads 71 | ^^^^^^^^^^^^ 72 | .. automodule:: mmrotate.models.dense_heads 73 | :members: 74 | 75 | roi_heads 76 | ^^^^^^^^^^ 77 | .. automodule:: mmrotate.models.roi_heads 78 | :members: 79 | 80 | losses 81 | ^^^^^^^^^^ 82 | .. automodule:: mmrotate.models.losses 83 | :members: 84 | 85 | utils 86 | ^^^^^^^^^^ 87 | .. automodule:: mmrotate.models.utils 88 | :members: 89 | 90 | mmrotate.utils 91 | -------------- 92 | .. automodule:: mmrotate.utils 93 | :members: 94 | -------------------------------------------------------------------------------- /docs/en/index.rst: -------------------------------------------------------------------------------- 1 | Welcome to MMRotate's documentation! 2 | ======================================= 3 | 4 | .. toctree:: 5 | :maxdepth: 1 6 | :caption: Get Started 7 | 8 | overview.md 9 | get_started.md 10 | 11 | .. toctree:: 12 | :maxdepth: 2 13 | :caption: User Guides 14 | 15 | user_guides/index.rst 16 | 17 | .. toctree:: 18 | :maxdepth: 2 19 | :caption: Advanced Guides 20 | 21 | advanced_guides/index.rst 22 | 23 | .. toctree:: 24 | :maxdepth: 1 25 | :caption: Migration 26 | 27 | migration.md 28 | 29 | .. toctree:: 30 | :maxdepth: 1 31 | :caption: API Reference 32 | 33 | api.rst 34 | 35 | .. toctree:: 36 | :maxdepth: 1 37 | :caption: Model Zoo 38 | 39 | model_zoo.md 40 | 41 | .. toctree:: 42 | :maxdepth: 1 43 | :caption: Notes 44 | 45 | notes/contribution_guide.md 46 | notes/projects.md 47 | notes/changelog.md 48 | notes/changelog_v0.x.md 49 | notes/faq.md 50 | 51 | .. toctree:: 52 | :caption: Switch Language 53 | 54 | switch_language.md 55 | 56 | 57 | 58 | Indices and tables 59 | ================== 60 | 61 | * :ref:`genindex` 62 | * :ref:`search` 63 | -------------------------------------------------------------------------------- /docs/en/make.bat: -------------------------------------------------------------------------------- 1 | @ECHO OFF 2 | 3 | pushd %~dp0 4 | 5 | REM Command file for Sphinx documentation 6 | 7 | if "%SPHINXBUILD%" == "" ( 8 | set SPHINXBUILD=sphinx-build 9 | ) 10 | set SOURCEDIR=. 11 | set BUILDDIR=_build 12 | 13 | if "%1" == "" goto help 14 | 15 | %SPHINXBUILD% >NUL 2>NUL 16 | if errorlevel 9009 ( 17 | echo. 18 | echo.The 'sphinx-build' command was not found. Make sure you have Sphinx 19 | echo.installed, then set the SPHINXBUILD environment variable to point 20 | echo.to the full path of the 'sphinx-build' executable. Alternatively you 21 | echo.may add the Sphinx directory to PATH. 22 | echo. 23 | echo.If you don't have Sphinx installed, grab it from 24 | echo.http://sphinx-doc.org/ 25 | exit /b 1 26 | ) 27 | 28 | %SPHINXBUILD% -M %1 %SOURCEDIR% %BUILDDIR% %SPHINXOPTS% %O% 29 | goto end 30 | 31 | :help 32 | %SPHINXBUILD% -M help %SOURCEDIR% %BUILDDIR% %SPHINXOPTS% %O% 33 | 34 | :end 35 | popd 36 | -------------------------------------------------------------------------------- /docs/en/migration.md: -------------------------------------------------------------------------------- 1 | # Migration 2 | -------------------------------------------------------------------------------- /docs/en/notes/projects.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/notes/projects.md -------------------------------------------------------------------------------- /docs/en/stat.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/stat.py -------------------------------------------------------------------------------- /docs/en/switch_language.md: -------------------------------------------------------------------------------- 1 | ## English 2 | 3 | ## 简体中文 4 | -------------------------------------------------------------------------------- /docs/en/user_guides/dataset_prepare.md: -------------------------------------------------------------------------------- 1 | # Dataset Preparation (To be updated) 2 | 3 | Please refer to [data preparation](https://github.com/open-mmlab/mmrotate/tree/main/tools/data) for dataset preparation. 4 | -------------------------------------------------------------------------------- /docs/en/user_guides/deploy.md: -------------------------------------------------------------------------------- 1 | # Model Deployment (To be updated) 2 | 3 | MMRotate 1.x fully relies on [MMDeploy](https://mmdeploy.readthedocs.io/) to deploy models. 4 | Please stay tuned and this document will be update soon. 5 | -------------------------------------------------------------------------------- /docs/en/user_guides/index.rst: -------------------------------------------------------------------------------- 1 | Train & Test 2 | ************** 3 | 4 | MMRotate provides dozens of pretrained detection models in `Model Zoo `_, 5 | and supports multiple standard datasets, including DOTA, HRSC2016, SSDD, HRSID, etc. This note will show how to perform common tasks on these existing models and standard datasets: 6 | 7 | 8 | .. toctree:: 9 | :maxdepth: 1 10 | 11 | config.md 12 | inference.md 13 | dataset_prepare.md 14 | train_test.md 15 | test_results_submission.md 16 | 17 | 18 | Useful Tools 19 | ************ 20 | 21 | .. toctree:: 22 | :maxdepth: 1 23 | 24 | useful_tools.md 25 | visualization.md 26 | deploy.md 27 | label_studio.md 28 | -------------------------------------------------------------------------------- /docs/en/user_guides/inference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/user_guides/inference.md -------------------------------------------------------------------------------- /docs/en/user_guides/test_results_submission.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/user_guides/test_results_submission.md -------------------------------------------------------------------------------- /docs/en/user_guides/visualization.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/en/user_guides/visualization.md -------------------------------------------------------------------------------- /docs/zh_cn/_static/css/readthedocs.css: -------------------------------------------------------------------------------- 1 | .header-logo { 2 | background-image: url("../image/mmrotate-logo.png"); 3 | background-size: 160px 40px; 4 | height: 40px; 5 | width: 160px; 6 | } 7 | -------------------------------------------------------------------------------- /docs/zh_cn/_static/image/mmrotate-logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/_static/image/mmrotate-logo.png -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/conventions.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/conventions.md -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/customize_losses.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/customize_losses.md -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/customize_transforms.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/customize_transforms.md -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/data_flow.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/data_flow.md -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/datasets.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/datasets.md -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/engine.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/engine.md -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/evaluation.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/evaluation.md -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/how_to.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/how_to.md -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/index.rst: -------------------------------------------------------------------------------- 1 | 基础概念 2 | *************** 3 | 4 | .. toctree:: 5 | :maxdepth: 1 6 | 7 | data_flow.md 8 | structures.md 9 | models.md 10 | datasets.md 11 | transforms.md 12 | evaluation.md 13 | engine.md 14 | conventions.md 15 | 16 | 组件定制 17 | ************************ 18 | 19 | .. toctree:: 20 | :maxdepth: 1 21 | 22 | customize_models.md 23 | customize_losses.md 24 | customize_dataset.md 25 | customize_transforms.md 26 | customize_runtime.md 27 | 28 | How to 29 | ************************ 30 | 31 | .. toctree:: 32 | :maxdepth: 1 33 | 34 | how_to.md 35 | -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/models.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/models.md -------------------------------------------------------------------------------- /docs/zh_cn/advanced_guides/transforms.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/advanced_guides/transforms.md -------------------------------------------------------------------------------- /docs/zh_cn/api.rst: -------------------------------------------------------------------------------- 1 | mmrotate.apis 2 | -------------- 3 | .. automodule:: mmrotate.apis 4 | :members: 5 | 6 | mmrotate.core 7 | -------------- 8 | 9 | anchor 10 | ^^^^^^^^^^ 11 | .. automodule:: mmrotate.core.anchor 12 | :members: 13 | 14 | bbox 15 | ^^^^^^^^^^ 16 | .. automodule:: mmrotate.core.bbox 17 | :members: 18 | 19 | patch 20 | ^^^^^^^^^^ 21 | .. automodule:: mmrotate.core.patch 22 | :members: 23 | 24 | evaluation 25 | ^^^^^^^^^^ 26 | .. automodule:: mmrotate.core.evaluation 27 | :members: 28 | 29 | post_processing 30 | ^^^^^^^^^^^^^^^ 31 | .. automodule:: mmrotate.core.post_processing 32 | :members: 33 | 34 | visualization 35 | ^^^^^^^^^^^^^ 36 | .. automodule:: mmrotate.core.visualization 37 | :members: 38 | 39 | mmrotate.datasets 40 | -------------- 41 | 42 | datasets 43 | ^^^^^^^^^^ 44 | .. automodule:: mmrotate.datasets 45 | :members: 46 | 47 | pipelines 48 | ^^^^^^^^^^ 49 | .. automodule:: mmrotate.datasets.pipelines 50 | :members: 51 | 52 | mmrotate.models 53 | -------------- 54 | 55 | detectors 56 | ^^^^^^^^^^ 57 | .. automodule:: mmrotate.models.detectors 58 | :members: 59 | 60 | backbones 61 | ^^^^^^^^^^ 62 | .. automodule:: mmrotate.models.backbones 63 | :members: 64 | 65 | necks 66 | ^^^^^^^^^^^^ 67 | .. automodule:: mmrotate.models.necks 68 | :members: 69 | 70 | dense_heads 71 | ^^^^^^^^^^^^ 72 | .. automodule:: mmrotate.models.dense_heads 73 | :members: 74 | 75 | roi_heads 76 | ^^^^^^^^^^ 77 | .. automodule:: mmrotate.models.roi_heads 78 | :members: 79 | 80 | losses 81 | ^^^^^^^^^^ 82 | .. automodule:: mmrotate.models.losses 83 | :members: 84 | 85 | utils 86 | ^^^^^^^^^^ 87 | .. automodule:: mmrotate.models.utils 88 | :members: 89 | 90 | mmrotate.utils 91 | -------------- 92 | .. automodule:: mmrotate.utils 93 | :members: 94 | -------------------------------------------------------------------------------- /docs/zh_cn/index.rst: -------------------------------------------------------------------------------- 1 | Welcome to MMRotate's documentation! 2 | ======================================= 3 | 4 | .. toctree:: 5 | :maxdepth: 1 6 | :caption: 开始你的第一步 7 | 8 | overview.md 9 | get_started.md 10 | 11 | .. toctree:: 12 | :maxdepth: 2 13 | :caption: 使用指南 14 | 15 | user_guides/index.rst 16 | 17 | .. toctree:: 18 | :maxdepth: 2 19 | :caption: 进阶教程 20 | 21 | advanced_guides/index.rst 22 | 23 | .. toctree:: 24 | :maxdepth: 1 25 | :caption: 迁移指南 26 | 27 | migration.md 28 | 29 | .. toctree:: 30 | :maxdepth: 1 31 | :caption: 接口文档（英文） 32 | 33 | api.rst 34 | 35 | .. toctree:: 36 | :maxdepth: 1 37 | :caption: 模型仓库 38 | 39 | model_zoo.md 40 | 41 | .. toctree:: 42 | :maxdepth: 1 43 | :caption: 说明 44 | 45 | notes/contribution_guide.md 46 | notes/projects.md 47 | notes/changelog.md 48 | notes/changelog_v0.x.md 49 | notes/faq.md 50 | 51 | .. toctree:: 52 | :caption: 语言切换 53 | 54 | switch_language.md 55 | 56 | 57 | 58 | Indices and tables 59 | ================== 60 | 61 | * :ref:`genindex` 62 | * :ref:`search` 63 | -------------------------------------------------------------------------------- /docs/zh_cn/make.bat: -------------------------------------------------------------------------------- 1 | @ECHO OFF 2 | 3 | pushd %~dp0 4 | 5 | REM Command file for Sphinx documentation 6 | 7 | if "%SPHINXBUILD%" == "" ( 8 | set SPHINXBUILD=sphinx-build 9 | ) 10 | set SOURCEDIR=. 11 | set BUILDDIR=_build 12 | 13 | if "%1" == "" goto help 14 | 15 | %SPHINXBUILD% >NUL 2>NUL 16 | if errorlevel 9009 ( 17 | echo. 18 | echo.The 'sphinx-build' command was not found. Make sure you have Sphinx 19 | echo.installed, then set the SPHINXBUILD environment variable to point 20 | echo.to the full path of the 'sphinx-build' executable. Alternatively you 21 | echo.may add the Sphinx directory to PATH. 22 | echo. 23 | echo.If you don't have Sphinx installed, grab it from 24 | echo.http://sphinx-doc.org/ 25 | exit /b 1 26 | ) 27 | 28 | %SPHINXBUILD% -M %1 %SOURCEDIR% %BUILDDIR% %SPHINXOPTS% %O% 29 | goto end 30 | 31 | :help 32 | %SPHINXBUILD% -M help %SOURCEDIR% %BUILDDIR% %SPHINXOPTS% %O% 33 | 34 | :end 35 | popd 36 | -------------------------------------------------------------------------------- /docs/zh_cn/migration.md: -------------------------------------------------------------------------------- 1 | # Migration 2 | -------------------------------------------------------------------------------- /docs/zh_cn/notes/contribution_guide.md: -------------------------------------------------------------------------------- 1 | # 贡献指南（待更新） 2 | -------------------------------------------------------------------------------- /docs/zh_cn/notes/projects.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/notes/projects.md -------------------------------------------------------------------------------- /docs/zh_cn/stat.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/stat.py -------------------------------------------------------------------------------- /docs/zh_cn/switch_language.md: -------------------------------------------------------------------------------- 1 | ## English 2 | 3 | ## 简体中文 4 | -------------------------------------------------------------------------------- /docs/zh_cn/user_guides/dataset_prepare.md: -------------------------------------------------------------------------------- 1 | # 准备数据集 (待更新) 2 | 3 | 具体的细节可以参考 [准备数据](https://github.com/open-mmlab/mmrotate/tree/main/tools/data) 下载并组织数据集。 4 | -------------------------------------------------------------------------------- /docs/zh_cn/user_guides/deploy.md: -------------------------------------------------------------------------------- 1 | # 模型部署 (待更新) 2 | 3 | MMRotate 1.x 完全基于 [MMDeploy](https://mmdeploy.readthedocs.io/) 來部署模型。 我们将在下一个版本完善这个文档。 4 | -------------------------------------------------------------------------------- /docs/zh_cn/user_guides/index.rst: -------------------------------------------------------------------------------- 1 | 训练 & 测试 2 | ************** 3 | 4 | MMRotate 提供了几十个预训练模型 `Model Zoo `_, 5 | 并且支持了多个标准数据集, 包括 DOTA, HRSC2016, SSDD, HRSID 等。本说明将展示如何在这些现有模型和标准数据集上执行常见任务： 6 | 7 | 8 | .. toctree:: 9 | :maxdepth: 1 10 | 11 | config.md 12 | inference.md 13 | dataset_prepare.md 14 | train_test.md 15 | test_results_submission.md 16 | 17 | 18 | 实用工具 19 | ************ 20 | 21 | .. toctree:: 22 | :maxdepth: 1 23 | 24 | useful_tools.md 25 | visualization.md 26 | deploy.md 27 | label_studio.md 28 | -------------------------------------------------------------------------------- /docs/zh_cn/user_guides/inference.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/user_guides/inference.md -------------------------------------------------------------------------------- /docs/zh_cn/user_guides/test_results_submission.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/user_guides/test_results_submission.md -------------------------------------------------------------------------------- /docs/zh_cn/user_guides/visualization.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/docs/zh_cn/user_guides/visualization.md -------------------------------------------------------------------------------- /mmrotate/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | import mmcv 3 | import mmdet 4 | import mmengine 5 | from mmengine.utils import digit_version 6 | 7 | from .version import __version__, short_version 8 | 9 | mmcv_minimum_version = '2.0.0rc4' 10 | mmcv_maximum_version = '2.1.0' 11 | mmcv_version = digit_version(mmcv.__version__) 12 | 13 | assert (mmcv_version >= digit_version(mmcv_minimum_version) 14 | and mmcv_version <= digit_version(mmcv_maximum_version)), \ 15 | f'MMCV {mmcv.__version__} is incompatible with MMRotate {__version__}. ' \ 16 | f'Please use MMCV >= {mmcv_minimum_version}, ' \ 17 | f'<= {mmcv_maximum_version} instead.' 18 | 19 | mmengine_minimum_version = '0.6.0' 20 | mmengine_maximum_version = '1.0.0' 21 | mmengine_version = digit_version(mmengine.__version__) 22 | 23 | assert (mmengine_version >= digit_version(mmengine_minimum_version) 24 | and mmengine_version < digit_version(mmengine_maximum_version)), \ 25 | f'MMEngine=={mmengine.__version__} is used but incompatible. ' \ 26 | f'Please install mmengine>={mmengine_minimum_version}, ' \ 27 | f'<{mmengine_maximum_version}.' 28 | 29 | mmdet_minimum_version = '3.0.0rc6' 30 | mmdet_maximum_version = '3.3.0' 31 | mmdet_version = digit_version(mmdet.__version__) 32 | 33 | assert (mmdet_version >= digit_version(mmdet_minimum_version) 34 | and mmdet_version <= digit_version(mmdet_maximum_version)), \ 35 | f'MMDetection {mmdet.__version__} is incompatible ' \ 36 | f'with MMRotate {__version__}. ' \ 37 | f'Please use MMDetection >= {mmdet_minimum_version}, ' \ 38 | f'< {mmdet_maximum_version} instead.' 39 | 40 | __all__ = ['__version__', 'short_version', 'digit_version'] 41 | -------------------------------------------------------------------------------- /mmrotate/apis/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .inference import inference_detector_by_patches 3 | 4 | __all__ = ['inference_detector_by_patches'] 5 | -------------------------------------------------------------------------------- /mmrotate/datasets/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .dior import DIORDataset # noqa: F401, F403 3 | from .dota import DOTAv2Dataset # noqa: F401, F403 4 | from .dota import DOTADataset, DOTAv15Dataset 5 | from .hrsc import HRSCDataset # noqa: F401, F403 6 | from .transforms import * # noqa: F401, F403 7 | from .nwpu45 import NWPU45Dataset 8 | from .instance_dataset import InstanceDataset 9 | 10 | __all__ = [ 11 | 'DOTADataset', 'DOTAv15Dataset', 'DOTAv2Dataset', 'HRSCDataset', 12 | 'DIORDataset', 'NWPU45Dataset', 'InstanceDataset' 13 | ] 14 | -------------------------------------------------------------------------------- /mmrotate/datasets/transforms/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .loading import LoadPatchFromNDArray, LoadEmptyAnnotations 3 | from .transforms import (ConvertBoxType, ConvertMask2BoxType, 4 | RandomChoiceRotate, RandomRotate, Rotate) 5 | 6 | __all__ = [ 7 | 'LoadPatchFromNDArray', 'Rotate', 'RandomRotate', 'RandomChoiceRotate', 8 | 'ConvertBoxType', 'ConvertMask2BoxType', 'LoadEmptyAnnotations' 9 | ] 10 | -------------------------------------------------------------------------------- /mmrotate/engine/__init__.py: -------------------------------------------------------------------------------- 1 | from .hooks import * -------------------------------------------------------------------------------- /mmrotate/engine/hooks/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .mean_teacher_hook import MeanTeacherHook 3 | 4 | __all__ = ['MeanTeacherHook'] -------------------------------------------------------------------------------- /mmrotate/evaluation/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .functional import * # noqa: F401,F403 3 | from .metrics import * # noqa: F401,F403 4 | -------------------------------------------------------------------------------- /mmrotate/evaluation/functional/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .mean_ap import eval_rbbox_map 3 | 4 | __all__ = ['eval_rbbox_map'] 5 | -------------------------------------------------------------------------------- /mmrotate/evaluation/metrics/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .dota_metric import DOTAMetric 3 | from .rotated_coco_metric import RotatedCocoMetric 4 | 5 | __all__ = ['DOTAMetric', 'RotatedCocoMetric'] 6 | -------------------------------------------------------------------------------- /mmrotate/models/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .backbones import * # noqa: F401, F403 3 | from .dense_heads import * # noqa: F401, F403 4 | from .detectors import * # noqa: F401, F403 5 | from .layers import * # noqa: F401, F403 6 | from .losses import * # noqa: F401, F403 7 | from .necks import * # noqa: F401, F403 8 | from .roi_heads import * # noqa: F401, F403 9 | from .task_modules import * # noqa: F401,F403 10 | from .utils import * # noqa: F401, F403 11 | -------------------------------------------------------------------------------- /mmrotate/models/backbones/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .re_resnet import ReResNet 3 | 4 | __all__ = ['ReResNet'] 5 | -------------------------------------------------------------------------------- /mmrotate/models/dense_heads/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .angle_branch_retina_head import AngleBranchRetinaHead 3 | from .cfa_head import CFAHead 4 | from .h2rbox_head import H2RBoxHead 5 | from .h2rbox_v2_head import H2RBoxV2Head 6 | from .oriented_reppoints_head import OrientedRepPointsHead 7 | from .oriented_rpn_head import OrientedRPNHead 8 | from .r3_head import R3Head, R3RefineHead 9 | from .rotated_atss_head import RotatedATSSHead 10 | from .rotated_fcos_head import RotatedFCOSHead 11 | from .rotated_reppoints_head import RotatedRepPointsHead 12 | from .rotated_retina_head import RotatedRetinaHead 13 | from .rotated_rtmdet_head import RotatedRTMDetHead, RotatedRTMDetSepBNHead 14 | from .s2a_head import S2AHead, S2ARefineHead 15 | from .sam_reppoints_head import SAMRepPointsHead 16 | from .rhino_align_head import RHINOAlignHead 17 | from .rhino_head import RHINOHead 18 | from .rhino_ph_head import RHINOPositiveHungarianHead 19 | from .rhino_phc_head import RHINOPositiveHungarianClassificationHead 20 | from .rotated_conditional_detr_head import RotatedConditionalDETRHead 21 | from .rotated_dab_detr_head import RotatedDABDETRHead 22 | from .rotated_deformable_detr_head import RotatedDeformableDETRHead 23 | from .rotated_detr_head import RotatedDETRHead 24 | 25 | 26 | __all__ = [ 27 | 'RotatedRetinaHead', 'OrientedRPNHead', 'RotatedRepPointsHead', 28 | 'SAMRepPointsHead', 'AngleBranchRetinaHead', 'RotatedATSSHead', 29 | 'RotatedFCOSHead', 'OrientedRepPointsHead', 'R3Head', 'R3RefineHead', 30 | 'S2AHead', 'S2ARefineHead', 'CFAHead', 'H2RBoxHead', 'H2RBoxV2Head', 31 | 'RotatedRTMDetHead', 'RotatedRTMDetSepBNHead', 'RotatedDETRHead', 32 | 'RotatedDeformableDETRHead', 'RotatedConditionalDETRHead', 33 | 'RotatedDABDETRHead', 'RHINOHead', 'RHINOPositiveHungarianHead', 34 | 'RHINOPositiveHungarianClassificationHead', 'RHINOAlignHead' 35 | ] 36 | -------------------------------------------------------------------------------- /mmrotate/models/dense_heads/rotated_conditional_detr_head.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) SI Analytics. All rights reserved. 2 | # Licensed under the CC BY-NC 4.0 License. See LICENSE file in the project root for full license information. 3 | # 4 | # Copyright (c) OpenMMLab. All rights reserved. 5 | # Licensed under the Apache License, Version 2.0. See LICENSE file in the mmrotate repository for full license information. 6 | from mmdet.models.dense_heads import ConditionalDETRHead 7 | 8 | from mmrotate.registry import MODELS 9 | from .rotated_detr_head import RotatedDETRHead 10 | 11 | 12 | @MODELS.register_module() 13 | class RotatedConditionalDETRHead(ConditionalDETRHead, RotatedDETRHead): 14 | """Rotated version of Head of Conditional DETR. 15 | 16 | Methods are inherited as follows. 17 | ConditionalDETRHead 18 | - init_weights 19 | - forward 20 | - loss 21 | - loss_and_predict 22 | - predict 23 | RotatedConditionalDETRHead 24 | - other methods 25 | """ 26 | pass 27 | -------------------------------------------------------------------------------- /mmrotate/models/dense_heads/rotated_dab_detr_head.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) SI Analytics. All rights reserved. 2 | # Licensed under the CC BY-NC 4.0 License. See LICENSE file in the project root for full license information. 3 | # 4 | # Copyright (c) OpenMMLab. All rights reserved. 5 | # Licensed under the Apache License, Version 2.0. See LICENSE file in the mmrotate repository for full license information. 6 | from mmcv.cnn import Linear 7 | from mmdet.models.dense_heads import DABDETRHead 8 | from mmdet.models.layers import MLP 9 | 10 | from mmrotate.registry import MODELS 11 | from .rotated_conditional_detr_head import RotatedConditionalDETRHead 12 | 13 | 14 | @MODELS.register_module() 15 | class RotatedDABDETRHead(DABDETRHead, RotatedConditionalDETRHead): 16 | """Rotated version of Head of DAB-DETR. 17 | 18 | Methods are inherited as follows. 19 | DABDETRHead 20 | - init_weights 21 | - forward 22 | - predict 23 | RotatedConditionalDETRHead 24 | - other methods 25 | """ 26 | 27 | def _init_layers(self) -> None: 28 | """Initialize layers of the transformer head.""" 29 | # cls branch 30 | self.fc_cls = Linear(self.embed_dims, self.cls_out_channels) 31 | # reg branch 32 | self.fc_reg = MLP(self.embed_dims, self.embed_dims, self.reg_dim, 3) 33 | -------------------------------------------------------------------------------- /mmrotate/models/detectors/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .h2rbox import H2RBoxDetector 3 | from .h2rbox_v2 import H2RBoxV2Detector 4 | from .refine_single_stage import RefineSingleStageDetector 5 | from .semi_base import RotatedSemiBaseDetector 6 | from .rotated_soft_teacher import RotatedSoftTeacher 7 | from .rhino import RHINO 8 | from .rotated_dab_detr import RotatedDABDETR 9 | from .rotated_deformable_detr import RotatedDeformableDETR 10 | 11 | __all__ = ['RefineSingleStageDetector', 'H2RBoxDetector', 'H2RBoxV2Detector', 12 | 'RotatedSemiBaseDetector', 'RotatedSoftTeacher', 13 | 'RotatedDABDETR', 'RotatedDeformableDETR', 'RHINO'] 14 | -------------------------------------------------------------------------------- /mmrotate/models/detectors/rotated_dab_detr.py: -------------------------------------------------------------------------------- 1 | # Ref: https://github.com/SIAnalytics/RHINO 2 | 3 | # Copyright (c) SI Analytics. All rights reserved. 4 | # Licensed under the CC BY-NC 4.0 License. See LICENSE file in the project root for full license information. 5 | # 6 | # Copyright (c) OpenMMLab. All rights reserved. 7 | # Licensed under the Apache License, Version 2.0. See LICENSE file in the mmrotate repository for full license information. 8 | from mmdet.models.detectors.dab_detr import DABDETR 9 | from mmdet.models.layers import SinePositionalEncoding 10 | from mmdet.models.layers.transformer import DABDetrTransformerEncoder 11 | from torch import nn 12 | 13 | from mmrotate.registry import MODELS 14 | from ..layers import RotatedDABDetrTransformerDecoder 15 | 16 | 17 | @MODELS.register_module() 18 | class RotatedDABDETR(DABDETR): 19 | r"""Angle refine version of DAB-DETR: 20 | 21 | Code is modified from the mmdet. 22 | """ 23 | 24 | def _init_layers(self) -> None: 25 | """Initialize decoder as RotatedDABDetrTransformerDecoder for angle 26 | refine.""" 27 | self.positional_encoding = SinePositionalEncoding( 28 | **self.positional_encoding) 29 | self.encoder = DABDetrTransformerEncoder(**self.encoder) 30 | self.decoder = RotatedDABDetrTransformerDecoder(**self.decoder) 31 | self.embed_dims = self.encoder.embed_dims 32 | self.query_dim = self.decoder.query_dim 33 | self.query_embedding = nn.Embedding(self.num_queries, self.query_dim) 34 | if self.num_patterns > 0: 35 | self.patterns = nn.Embedding(self.num_patterns, self.embed_dims) 36 | 37 | num_feats = self.positional_encoding.num_feats 38 | assert num_feats * 2 == self.embed_dims, \ 39 | f'embed_dims should be exactly 2 times of num_feats. ' \ 40 | f'Found {self.embed_dims} and {num_feats}.' 41 | -------------------------------------------------------------------------------- /mmrotate/models/layers/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .align import FRM, AlignConv, DCNAlignModule, PseudoAlignModule 3 | from .transformer import (RhinoTransformerDecoder, RhinoTransformerDecoderV2, 4 | RhinoTransformerDecoderV4, RotatedCdnQueryGenerator, 5 | RotatedDABDetrTransformerDecoder, 6 | RotatedDeformableDetrTransformerDecoder, 7 | coordinate_to_encoding, RotatedMultiScaleDeformableAttention) 8 | 9 | __all__ = ['FRM', 'AlignConv', 'DCNAlignModule', 'PseudoAlignModule', 10 | 'coordinate_to_encoding', 'RotatedDABDetrTransformerDecoder', 11 | 'RotatedDeformableDetrTransformerDecoder', 'RhinoTransformerDecoder', 12 | 'RotatedCdnQueryGenerator', 'RhinoTransformerDecoderV2', 13 | 'RhinoTransformerDecoderV4', 'RotatedMultiScaleDeformableAttention'] 14 | -------------------------------------------------------------------------------- /mmrotate/models/layers/transformer/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) SI Analytics. All rights reserved. 2 | # Licensed under the CC BY-NC 4.0 License. See LICENSE file in the project root for full license information. 3 | # 4 | # Copyright (c) OpenMMLab. All rights reserved. 5 | # Licensed under the Apache License, Version 2.0. See LICENSE file in the mmrotate repository for full license information. 6 | from .rhino_layers import RhinoTransformerDecoder, RotatedCdnQueryGenerator 7 | from .rhino_layers_v2 import (RhinoTransformerDecoderV2, 8 | RhinoTransformerDecoderV4) 9 | from .rotated_dab_detr_layers import RotatedDABDetrTransformerDecoder 10 | from .rotated_deformable_detr_layers import \ 11 | RotatedDeformableDetrTransformerDecoder 12 | from .utils import coordinate_to_encoding 13 | from .rotated_attention import RotatedMultiScaleDeformableAttention 14 | 15 | __all__ = [ 16 | 'coordinate_to_encoding', 17 | 'RotatedDABDetrTransformerDecoder', 18 | 'RotatedDeformableDetrTransformerDecoder', 19 | 'RhinoTransformerDecoder', 20 | 'RotatedCdnQueryGenerator', 21 | 'RhinoTransformerDecoderV2', 22 | 'RhinoTransformerDecoderV4', 23 | 'RotatedMultiScaleDeformableAttention' 24 | ] 25 | -------------------------------------------------------------------------------- /mmrotate/models/losses/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .convex_giou_loss import BCConvexGIoULoss, ConvexGIoULoss 3 | from .gaussian_dist_loss import GDLoss 4 | from .gaussian_dist_loss_v1 import GDLoss_v1 5 | from .h2rbox_consistency_loss import H2RBoxConsistencyLoss 6 | from .h2rbox_v2_consistency_loss import H2RBoxV2ConsistencyLoss 7 | from .kf_iou_loss import KFLoss 8 | from .rotated_iou_loss import RotatedIoULoss 9 | from .smooth_focal_loss import SmoothFocalLoss 10 | from .spatial_border_loss import SpatialBorderLoss 11 | 12 | __all__ = [ 13 | 'GDLoss', 'GDLoss_v1', 'KFLoss', 'ConvexGIoULoss', 'BCConvexGIoULoss', 14 | 'SmoothFocalLoss', 'RotatedIoULoss', 'SpatialBorderLoss', 15 | 'H2RBoxConsistencyLoss', 'H2RBoxV2ConsistencyLoss' 16 | ] 17 | -------------------------------------------------------------------------------- /mmrotate/models/necks/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .re_fpn import ReFPN 3 | 4 | __all__ = ['ReFPN'] 5 | -------------------------------------------------------------------------------- /mmrotate/models/roi_heads/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .bbox_heads import RotatedShared2FCBBoxHead 3 | from .gv_ratio_roi_head import GVRatioRoIHead 4 | from .roi_extractors import RotatedSingleRoIExtractor 5 | 6 | __all__ = [ 7 | 'RotatedShared2FCBBoxHead', 'RotatedSingleRoIExtractor', 'GVRatioRoIHead' 8 | ] 9 | -------------------------------------------------------------------------------- /mmrotate/models/roi_heads/bbox_heads/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .convfc_rbbox_head import RotatedShared2FCBBoxHead 3 | from .gv_bbox_head import GVBBoxHead 4 | 5 | __all__ = ['RotatedShared2FCBBoxHead', 'GVBBoxHead'] 6 | -------------------------------------------------------------------------------- /mmrotate/models/roi_heads/roi_extractors/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .rotate_single_level_roi_extractor import RotatedSingleRoIExtractor 3 | 4 | __all__ = ['RotatedSingleRoIExtractor'] 5 | -------------------------------------------------------------------------------- /mmrotate/models/task_modules/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .assigners import * # noqa: F401,F403 3 | from .coders import * # noqa: F401,F403 4 | from .prior_generators import * # noqa: F401,F403 5 | -------------------------------------------------------------------------------- /mmrotate/models/task_modules/assigners/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .convex_assigner import ConvexAssigner 3 | from .max_convex_iou_assigner import MaxConvexIoUAssigner 4 | from .rotate_iou2d_calculator import (FakeRBboxOverlaps2D, 5 | QBbox2HBboxOverlaps2D, 6 | RBbox2HBboxOverlaps2D, RBboxOverlaps2D) 7 | from .rotated_atss_assigner import RotatedATSSAssigner 8 | from .sas_assigner import SASAssigner 9 | from .dn_group_hungarian_assigner import DNGroupHungarianAssigner 10 | from .match_cost import CenterL1Cost, GDCost, RBoxL1Cost, RotatedIoUCost 11 | 12 | __all__ = [ 13 | 'ConvexAssigner', 'MaxConvexIoUAssigner', 'SASAssigner', 14 | 'RotatedATSSAssigner', 'RBboxOverlaps2D', 'FakeRBboxOverlaps2D', 15 | 'RBbox2HBboxOverlaps2D', 'QBbox2HBboxOverlaps2D', 'RBoxL1Cost', 'GDCost', 16 | 'RotatedIoUCost', 'CenterL1Cost', 'DNGroupHungarianAssigner' 17 | ] 18 | -------------------------------------------------------------------------------- /mmrotate/models/task_modules/coders/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .angle_coder import CSLCoder, PSCCoder, PseudoAngleCoder 3 | from .delta_midpointoffset_rbbox_coder import MidpointOffsetCoder 4 | from .delta_xywh_hbbox_coder import DeltaXYWHHBBoxCoder 5 | from .delta_xywh_qbbox_coder import DeltaXYWHQBBoxCoder 6 | from .delta_xywht_hbbox_coder import DeltaXYWHTHBBoxCoder 7 | from .delta_xywht_rbbox_coder import DeltaXYWHTRBBoxCoder 8 | from .distance_angle_point_coder import DistanceAnglePointCoder 9 | from .gliding_vertex_coder import GVFixCoder, GVRatioCoder 10 | 11 | __all__ = [ 12 | 'DeltaXYWHTRBBoxCoder', 'DeltaXYWHTHBBoxCoder', 'MidpointOffsetCoder', 13 | 'GVFixCoder', 'GVRatioCoder', 'CSLCoder', 'PSCCoder', 14 | 'DistanceAnglePointCoder', 'DeltaXYWHHBBoxCoder', 'DeltaXYWHQBBoxCoder', 15 | 'PseudoAngleCoder' 16 | ] 17 | -------------------------------------------------------------------------------- /mmrotate/models/task_modules/coders/delta_xywh_hbbox_coder.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from typing import Union 3 | 4 | from mmdet.models.task_modules.coders import DeltaXYWHBBoxCoder 5 | from mmdet.models.task_modules.coders.delta_xywh_bbox_coder import bbox2delta 6 | from mmdet.structures.bbox import HorizontalBoxes, get_box_tensor 7 | from torch import Tensor 8 | 9 | from mmrotate.registry import TASK_UTILS 10 | from mmrotate.structures.bbox import RotatedBoxes 11 | 12 | 13 | @TASK_UTILS.register_module() 14 | class DeltaXYWHHBBoxCoder(DeltaXYWHBBoxCoder): 15 | """Delta XYWH HBBox coder. 16 | 17 | This coder is almost the same as `DeltaXYWHBBoxCoder`. Besides the 18 | gt_bboxes of encode is :obj:`RotatedBoxes`. 19 | """ 20 | 21 | def encode(self, bboxes: Union[HorizontalBoxes, Tensor], 22 | gt_bboxes: Union[RotatedBoxes, Tensor]) -> Tensor: 23 | """Get box regression transformation deltas that can be used to 24 | transform the ``bboxes`` into the ``gt_bboxes``. 25 | 26 | Args: 27 | bboxes (:obj:`HorizontalBoxes` or Tensor): Source boxes, e.g., 28 | object proposals. 29 | gt_bboxes (:obj:`RotatedBoxes` or Tensor): Target of the 30 | transformation, e.g., ground-truth boxes. 31 | Returns: 32 | Tensor: Box transformation deltas 33 | """ 34 | 35 | assert bboxes.size(0) == gt_bboxes.size(0) 36 | assert bboxes.size(-1) == 4 37 | assert gt_bboxes.size(-1) == 5 38 | 39 | bboxes = get_box_tensor(bboxes) 40 | 41 | if not isinstance(gt_bboxes, RotatedBoxes): 42 | gt_bboxes = RotatedBoxes(gt_bboxes) 43 | gt_bboxes = gt_bboxes.convert_to('hbox').tensor 44 | 45 | encoded_bboxes = bbox2delta(bboxes, gt_bboxes, self.means, self.stds) 46 | return encoded_bboxes 47 | -------------------------------------------------------------------------------- /mmrotate/models/task_modules/coders/delta_xywh_qbbox_coder.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from typing import Union 3 | 4 | from mmdet.models.task_modules.coders import DeltaXYWHBBoxCoder 5 | from mmdet.models.task_modules.coders.delta_xywh_bbox_coder import bbox2delta 6 | from mmdet.structures.bbox import HorizontalBoxes, get_box_tensor 7 | from torch import Tensor 8 | 9 | from mmrotate.registry import TASK_UTILS 10 | from mmrotate.structures.bbox import QuadriBoxes 11 | 12 | 13 | @TASK_UTILS.register_module() 14 | class DeltaXYWHQBBoxCoder(DeltaXYWHBBoxCoder): 15 | """Delta XYWH QBBox coder. 16 | 17 | This coder is almost the same as `DeltaXYWHBBoxCoder`. Besides the 18 | gt_bboxes of encode is :obj:`QuadriBoxes`. 19 | """ 20 | 21 | def encode(self, bboxes: Union[HorizontalBoxes, Tensor], 22 | gt_bboxes: Union[QuadriBoxes, Tensor]) -> Tensor: 23 | """Get box regression transformation deltas that can be used to 24 | transform the ``bboxes`` into the ``gt_bboxes``. 25 | 26 | Args: 27 | bboxes (:obj:`HorizontalBoxes` or Tensor): Source boxes, e.g., 28 | object proposals. 29 | gt_bboxes (:obj:`QuadriBoxes` or Tensor): Target of the 30 | transformation, e.g., ground-truth boxes. 31 | Returns: 32 | Tensor: Box transformation deltas 33 | """ 34 | 35 | assert bboxes.size(0) == gt_bboxes.size(0) 36 | assert bboxes.size(-1) == 4 37 | assert gt_bboxes.size(-1) == 8 38 | 39 | bboxes = get_box_tensor(bboxes) 40 | 41 | if not isinstance(gt_bboxes, QuadriBoxes): 42 | gt_bboxes = QuadriBoxes(gt_bboxes) 43 | gt_bboxes = gt_bboxes.convert_to('hbox').tensor 44 | 45 | encoded_bboxes = bbox2delta(bboxes, gt_bboxes, self.means, self.stds) 46 | return encoded_bboxes 47 | -------------------------------------------------------------------------------- /mmrotate/models/task_modules/prior_generators/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .anchor_generator import (FakeRotatedAnchorGenerator, 3 | PseudoRotatedAnchorGenerator) 4 | 5 | __all__ = ['PseudoRotatedAnchorGenerator', 'FakeRotatedAnchorGenerator'] 6 | -------------------------------------------------------------------------------- /mmrotate/models/utils/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .misc import (convex_overlaps, get_num_level_anchors_inside, 3 | levels_to_images, points_center_pts, filter_gt_instances, 4 | _filter_rpn_results_by_score) 5 | from .orconv import ORConv2d 6 | from .ripool import RotationInvariantPooling 7 | 8 | __all__ = [ 9 | 'ORConv2d', 'RotationInvariantPooling', 'get_num_level_anchors_inside', 10 | 'points_center_pts', 'levels_to_images', 'convex_overlaps', 'filter_gt_instances', 11 | '_filter_rpn_results_by_score' 12 | ] 13 | -------------------------------------------------------------------------------- /mmrotate/models/utils/ripool.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from torch import nn 3 | 4 | 5 | class RotationInvariantPooling(nn.Module): 6 | """Rotating invariant pooling module. 7 | 8 | Args: 9 | nInputPlane (int): The number of Input plane. 10 | nOrientation (int, optional): The number of oriented channels. 11 | """ 12 | 13 | def __init__(self, nInputPlane, nOrientation=8): 14 | super(RotationInvariantPooling, self).__init__() 15 | self.nInputPlane = nInputPlane 16 | self.nOrientation = nOrientation 17 | 18 | def forward(self, x): 19 | """Forward function.""" 20 | N, c, h, w = x.size() 21 | x = x.view(N, -1, self.nOrientation, h, w) 22 | x, _ = x.max(dim=2, keepdim=False) 23 | return x 24 | -------------------------------------------------------------------------------- /mmrotate/structures/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .bbox import * # noqa: F401, F403 3 | -------------------------------------------------------------------------------- /mmrotate/structures/bbox/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .bbox_overlaps import fake_rbbox_overlaps, rbbox_overlaps 3 | from .box_converters import (hbox2qbox, hbox2rbox, qbox2hbox, qbox2rbox, 4 | rbox2hbox, rbox2qbox) 5 | from .quadri_boxes import QuadriBoxes 6 | from .rotated_boxes import RotatedBoxes 7 | from .transforms import (distance2obb, gaussian2bbox, gt2gaussian, norm_angle, 8 | rbox_project) 9 | 10 | __all__ = [ 11 | 'QuadriBoxes', 'RotatedBoxes', 'hbox2rbox', 'hbox2qbox', 'rbox2hbox', 12 | 'rbox2qbox', 'qbox2hbox', 'qbox2rbox', 'gaussian2bbox', 'gt2gaussian', 13 | 'norm_angle', 'rbbox_overlaps', 'fake_rbbox_overlaps', 'distance2obb', 14 | 'rbox_project' 15 | ] -------------------------------------------------------------------------------- /mmrotate/testing/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from ._utils import demo_mm_inputs, demo_mm_proposals, get_detector_cfg 3 | 4 | __all__ = ['get_detector_cfg', 'demo_mm_inputs', 'demo_mm_proposals'] 5 | -------------------------------------------------------------------------------- /mmrotate/utils/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .collect_env import collect_env 3 | from .misc import get_test_pipeline_cfg 4 | from .patch import get_multiscale_patch, merge_results_by_nms, slide_window 5 | from .setup_env import register_all_modules 6 | 7 | __all__ = [ 8 | 'collect_env', 'register_all_modules', 'get_test_pipeline_cfg', 9 | 'get_multiscale_patch', 'merge_results_by_nms', 'slide_window' 10 | ] 11 | -------------------------------------------------------------------------------- /mmrotate/utils/collect_env.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from mmengine.utils import get_git_hash 3 | from mmengine.utils.dl_utils import collect_env as collect_base_env 4 | 5 | import mmrotate 6 | 7 | 8 | def collect_env(): 9 | """Collect environment information.""" 10 | env_info = collect_base_env() 11 | env_info['MMRotate'] = ( 12 | mmrotate.__version__ + '+' + get_git_hash(digits=7)) 13 | return env_info 14 | 15 | 16 | if __name__ == '__main__': 17 | for name, val in collect_env().items(): 18 | print(f'{name}: {val}') 19 | -------------------------------------------------------------------------------- /mmrotate/utils/misc.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from typing import Union 3 | 4 | from mmengine.config import Config, ConfigDict 5 | 6 | 7 | def get_test_pipeline_cfg(cfg: Union[str, ConfigDict]) -> ConfigDict: 8 | """Get the test dataset pipeline from entire config. 9 | 10 | Args: 11 | cfg (str or :obj:`ConfigDict`): the entire config. Can be a config 12 | file or a ``ConfigDict``. 13 | 14 | Returns: 15 | :obj:`ConfigDict`: the config of test dataset. 16 | """ 17 | if isinstance(cfg, str): 18 | cfg = Config.fromfile(cfg) 19 | 20 | def _get_test_pipeline_cfg(dataset_cfg): 21 | if 'pipeline' in dataset_cfg: 22 | return dataset_cfg.pipeline 23 | # handle dataset wrapper 24 | elif 'dataset' in dataset_cfg: 25 | return _get_test_pipeline_cfg(dataset_cfg.dataset) 26 | # handle dataset wrappers like ConcatDataset 27 | elif 'datasets' in dataset_cfg: 28 | return _get_test_pipeline_cfg(dataset_cfg.datasets[0]) 29 | 30 | raise RuntimeError('Cannot find `pipeline` in `test_dataloader`') 31 | 32 | return _get_test_pipeline_cfg(cfg.test_dataloader.dataset) 33 | -------------------------------------------------------------------------------- /mmrotate/utils/patch/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .merge_results import merge_results_by_nms 3 | from .split import get_multiscale_patch, slide_window 4 | 5 | __all__ = ['merge_results_by_nms', 'get_multiscale_patch', 'slide_window'] 6 | -------------------------------------------------------------------------------- /mmrotate/utils/setup_env.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | import datetime 3 | import warnings 4 | 5 | from mmengine import DefaultScope 6 | 7 | 8 | def register_all_modules(init_default_scope: bool = True) -> None: 9 | """Register all modules in mmrotate into the registries. 10 | 11 | Args: 12 | init_default_scope (bool): Whether initialize the mmrotate default scope. 13 | When `init_default_scope=True`, the global default scope will be 14 | set to `mmrotate`, anmmrotate all registries will build modules from mmrotate's 15 | registry node. To understand more about the registry, please refer 16 | to https://github.com/open-mmlab/mmengine/blob/main/docs/en/tutorials/registry.md 17 | Defaults to True. 18 | """ # noqa 19 | import mmrotate.datasets # noqa: F401,F403 20 | import mmrotate.evaluation # noqa: F401,F403 21 | import mmrotate.models # noqa: F401,F403 22 | import mmrotate.visualization # noqa: F401,F403 23 | 24 | if init_default_scope: 25 | never_created = DefaultScope.get_current_instance() is None \ 26 | or not DefaultScope.check_instance_created('mmrotate') 27 | if never_created: 28 | DefaultScope.get_instance('mmrotate', scope_name='mmrotate') 29 | return 30 | current_scope = DefaultScope.get_current_instance() 31 | if current_scope.scope_name != 'mmrotate': 32 | warnings.warn('The current default scope ' 33 | f'"{current_scope.scope_name}" is not "mmrotate", ' 34 | '`register_all_modules` will force the current' 35 | 'default scope to be "mmrotate". If this is not ' 36 | 'expected, please set `init_default_scope=False`.') 37 | # avoid name conflict 38 | new_instance_name = f'mmrotate-{datetime.datetime.now()}' 39 | DefaultScope.get_instance(new_instance_name, scope_name='mmrotate') 40 | -------------------------------------------------------------------------------- /mmrotate/version.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | 3 | __version__ = '1.0.0rc1' 4 | short_version = __version__ 5 | -------------------------------------------------------------------------------- /mmrotate/visualization/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .local_visualizer import RotLocalVisualizer 3 | from .palette import get_palette 4 | 5 | __all__ = ['get_palette', 'RotLocalVisualizer'] 6 | -------------------------------------------------------------------------------- /mmrotate/visualization/palette.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from typing import List, Tuple, Union 3 | 4 | import mmcv 5 | import numpy as np 6 | from mmengine.utils import is_str 7 | 8 | 9 | def get_palette(palette: Union[List[tuple], str, tuple], 10 | num_classes: int) -> List[Tuple[int]]: 11 | """Get palette from various inputs. 12 | 13 | Args: 14 | palette (list[tuple] | str | tuple): palette inputs. 15 | num_classes (int): the number of classes. 16 | Returns: 17 | list[tuple[int]]: A list of color tuples. 18 | """ 19 | assert isinstance(num_classes, int) 20 | 21 | if isinstance(palette, list): 22 | dataset_palette = palette 23 | elif isinstance(palette, tuple): 24 | dataset_palette = [palette] * num_classes 25 | elif palette == 'random' or palette is None: 26 | state = np.random.get_state() 27 | # random color 28 | np.random.seed(42) 29 | palette = np.random.randint(0, 256, size=(num_classes, 3)) 30 | np.random.set_state(state) 31 | dataset_palette = [tuple(c) for c in palette] 32 | elif palette == 'dota': 33 | from mmrotate.datasets import DOTADataset 34 | dataset_palette = DOTADataset.METAINFO['palette'] 35 | elif palette == 'sar': 36 | from mmrotate.datasets import SARDataset 37 | dataset_palette = SARDataset.METAINFO['palette'] 38 | elif palette == 'hrsc': 39 | from mmrotate.datasets import HRSCDataset 40 | dataset_palette = HRSCDataset.METAINFO['palette'] 41 | elif is_str(palette): 42 | dataset_palette = [mmcv.color_val(palette)[::-1]] * num_classes 43 | else: 44 | raise TypeError(f'Invalid type for palette: {type(palette)}') 45 | 46 | assert len(dataset_palette) >= num_classes, \ 47 | 'The length of palette should not be less than `num_classes`.' 48 | return dataset_palette 49 | -------------------------------------------------------------------------------- /model-index.yml: -------------------------------------------------------------------------------- 1 | Import: 2 | - configs/cfa/metafile.yml 3 | - configs/convnext/metafile.yml 4 | - configs/csl/metafile.yml 5 | - configs/gliding_vertex/metafile.yml 6 | - configs/gwd/metafile.yml 7 | - configs/h2rbox/metafile.yml 8 | - configs/kfiou/metafile.yml 9 | - configs/kld/metafile.yml 10 | - configs/oriented_rcnn/metafile.yml 11 | - configs/oriented_reppoints/metafile.yml 12 | - configs/psc/metafile.yml 13 | - configs/r3det/metafile.yml 14 | - configs/redet/metafile.yml 15 | - configs/roi_trans/metafile.yml 16 | - configs/rotated_atss/metafile.yml 17 | - configs/rotated_faster_rcnn/metafile.yml 18 | - configs/rotated_fcos/metafile.yml 19 | - configs/rotated_reppoints/metafile.yml 20 | - configs/rotated_retinanet/metafile.yml 21 | - configs/rotated_rtmdet/metafile.yml 22 | - configs/s2anet/metafile.yml 23 | - configs/sasm_reppoints/metafile.yml 24 | -------------------------------------------------------------------------------- /projects/CastDetv2/castdet/__init__.py: -------------------------------------------------------------------------------- 1 | from .castdet import RotatedCastDet 2 | from .ovd_bbox_head import Shared2FCBBoxHeadZSD, Projection2 3 | from .pseudo_label_queue import PseudoQueue 4 | from .modified_resnet import ModifiedResNet2 5 | from .standard_roi_head2 import StandardRoIHead2 6 | 7 | __all__ = [ 8 | 'RotatedCastDet', 'Shared2FCBBoxHeadZSD', 'Projection2', 'PseudoQueue', 9 | 'ModifiedResNet2', 'StandardRoIHead2' 10 | ] -------------------------------------------------------------------------------- /projects/CastDetv2/configs/oriented-rcnn_r50-fpn_20k_visdronezsd_base-set.py: -------------------------------------------------------------------------------- 1 | _base_ = [ 2 | 'mmrotate::_base_/models/oriented-rcnn-le90_r50_fpn.py', 3 | 'mmrotate::_base_/default_runtime.py', 4 | 'mmrotate::_base_/datasets/visdronezsd.py' 5 | ] 6 | 7 | batch_size = 8 8 | num_workers = 2 9 | train_dataloader = dict( 10 | batch_size=batch_size, 11 | num_workers=num_workers, 12 | 13 | ) 14 | 15 | model = dict( 16 | roi_head = dict( 17 | bbox_head = dict(num_classes=16) 18 | ) 19 | ) 20 | 21 | # training schedule for 180k 22 | train_cfg = dict( 23 | type='IterBasedTrainLoop', max_iters=20000, val_interval=4000) 24 | val_cfg = dict(type='ValLoop') 25 | test_cfg = dict(type='TestLoop') 26 | 27 | # learning rate policy 28 | param_scheduler = [ 29 | dict( 30 | type='LinearLR', start_factor= 1.0 / 3, by_epoch=False, begin=0, end=500), 31 | dict( 32 | type='MultiStepLR', 33 | begin=0, 34 | end=20000, 35 | by_epoch=False, 36 | milestones=[16000, 18000], 37 | gamma=0.1) 38 | ] 39 | 40 | # optimizer 41 | optim_wrapper = dict( 42 | type='OptimWrapper', 43 | optimizer=dict(type='SGD', lr=0.005, momentum=0.9, weight_decay=0.0001), 44 | clip_grad=dict(max_norm=35, norm_type=2)) 45 | 46 | 47 | default_hooks = dict( 48 | logger=dict(type='LoggerHook', interval=20), 49 | checkpoint=dict(by_epoch=False, interval=4000, max_keep_ckpts=5)) 50 | log_processor = dict(by_epoch=False) 51 | 52 | visualizer = dict( 53 | vis_backends=[ 54 | dict(type='LocalVisBackend'), 55 | dict(type='TensorboardVisBackend') 56 | ]) 57 | 58 | # for debug 59 | # load_from = "work_dirs/soft-teacher_faster-rcnn_r50-caffe_fpn_80k_semi-dior/iter_10000.pth" -------------------------------------------------------------------------------- /projects/CastDetv2/resources/castdet.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/projects/CastDetv2/resources/castdet.png -------------------------------------------------------------------------------- /projects/CastDetv2/resources/remoteCLIP_embeddings_bgs_normalized.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/projects/CastDetv2/resources/remoteCLIP_embeddings_bgs_normalized.npy -------------------------------------------------------------------------------- /projects/CastDetv2/resources/remoteCLIP_embeddings_normalized.npy: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/projects/CastDetv2/resources/remoteCLIP_embeddings_normalized.npy -------------------------------------------------------------------------------- /projects/CastDetv2/resources/results_hbb.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/projects/CastDetv2/resources/results_hbb.png -------------------------------------------------------------------------------- /projects/CastDetv2/resources/results_obb.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/projects/CastDetv2/resources/results_obb.png -------------------------------------------------------------------------------- /projects/CastDetv2/resources/toolbox.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/projects/CastDetv2/resources/toolbox.png -------------------------------------------------------------------------------- /projects/CastDetv2/resources/vis_result.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/projects/CastDetv2/resources/vis_result.png -------------------------------------------------------------------------------- /projects/CastDetv2/run.sh: -------------------------------------------------------------------------------- 1 | DEVICES_ID=2 2 | exp="oriented-rcnn_r50-fpn_20k_visdronezsd_base-set" 3 | 4 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/train.py \ 5 | projects/CastDetv2/configs/$exp.py 6 | 7 | python projects/CastDetv2/tools/merge_weights.py \ 8 | --clip_path checkpoints/RemoteCLIP-RN50.pt \ 9 | --base_path work_dirs/$exp/iter_20000.pth \ 10 | --save_path work_dirs/$exp/merged_castdet_init_iter20k.pth \ 11 | --base_model faster-rcnn 12 | 13 | 14 | exp="visdrone_step2_castdet_12b_10k_oriented" 15 | 16 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/train.py \ 17 | projects/CastDetv2/configs/$exp.py 18 | 19 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/test.py \ 20 | projects/CastDetv2/configs/$exp.py \ 21 | work_dirs/$exp/iter_10000.pth \ 22 | --work-dir work_dirs/$exp/dior_test 23 | -------------------------------------------------------------------------------- /projects/CastDetv2/tools/merge_weights.py: -------------------------------------------------------------------------------- 1 | #!/usr/bin/python 2 | # -*- encoding: utf-8 -*- 3 | ''' 4 | @File : merge_weights.py 5 | @Version : 1.0 6 | @Time : 2024/04/20 00:00:16 7 | @E-mail : daodao123@sjtu.edu.cn 8 | @Introduction : None 9 | ''' 10 | 11 | import argparse 12 | import torch 13 | 14 | def merge_weights(clip_path, base_path, save_path, base_model='soft-teacher', target_model='castdet'): 15 | clip = torch.load(clip_path)#.state_dict() 16 | base = torch.load(base_path) 17 | save_dict = {} 18 | if target_model == 'castdet': 19 | for k, v in base['state_dict'].items(): 20 | if base_model == 'soft-teacher': 21 | save_dict[k] = v 22 | elif base_model == 'faster-rcnn': 23 | save_dict['teacher.' + k] = v 24 | save_dict['student.' + k] = v 25 | elif target_model == 'vild': 26 | save_dict = base['state_dict'] 27 | 28 | for k, v in clip.items(): 29 | if k.startswith('visual.'): 30 | save_dict[k] = v 31 | 32 | print(save_dict.keys()) 33 | torch.save(save_dict, save_path) 34 | 35 | def main(): 36 | parser = argparse.ArgumentParser(description="Merge weights from CLIP and a base detection model") 37 | parser.add_argument("--clip_path", type=str, required=True, help="Path to the CLIP model checkpoint") 38 | parser.add_argument("--base_path", type=str, required=True, help="Path to the base model checkpoint") 39 | parser.add_argument("--save_path", type=str, required=True, help="Path where the merged model will be saved") 40 | parser.add_argument("--base_model", type=str, default="soft-teacher", choices=["soft-teacher", "faster-rcnn"], help="Base model type: 'soft-teacher' or 'faster-rcnn'") 41 | parser.add_argument("--target_model", type=str, default="castdet", choices=["castdet", "vild"], help="Target model type: 'castdet' or 'vild'") 42 | 43 | args = parser.parse_args() 44 | merge_weights(args.clip_path, args.base_path, args.save_path, args.base_model) 45 | 46 | if __name__ == "__main__": 47 | main() 48 | -------------------------------------------------------------------------------- /projects/GLIP/glip/__init__.py: -------------------------------------------------------------------------------- 1 | from .atss_vlfusion_head import RotatedATSSVLFusionHead 2 | 3 | __all__ = [ 4 | 'RotatedATSSVLFusionHead' 5 | ] -------------------------------------------------------------------------------- /projects/GLIP/run.sh: -------------------------------------------------------------------------------- 1 | DEVICES_ID=3 2 | 3 | exp1="glip_atss_r50_a_fpn_dyhead_visdronezsd_base" 4 | 5 | # Step1: train base-detector 6 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/train.py \ 7 | projects/GLIP/configs/$exp1.py 8 | 9 | # Step2.1: pseudo-labeling 10 | exp2="glip_atss_r50_a_fpn_dyhead_visdronezsd_base_nwpu45_pseudo_labeling" 11 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/test.py \ 12 | projects/GLIP/configs/$exp2.py \ 13 | work_dirs/$exp1/iter_20000.pth 14 | 15 | # Step2.2: merge predictions 16 | python projects/GroundingDINO/tools/merge_ovdg_preds.py \ 17 | --ann_path data/NWPU-RESISC45/annotations/nwpu45_unlabeled_2.json \ 18 | --pred_path work_dirs/$exp2/nwpu45_pseudo_labeling_2.bbox.json \ 19 | --save_path work_dirs/$exp2/nwpu45_unlabeled_with_glip_pseudos_2.json 20 | 21 | cp work_dirs/$exp2/nwpu45_unlabeled_with_glip_pseudos_2.json data/NWPU-RESISC45/annotations/nwpu45_unlabeled_with_glip_pseudos_2.json 22 | 23 | # Step3: self-training 24 | exp3="glip_atss_r50_a_fpn_dyhead_visdronezsd_base_nwpu" 25 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/train.py \ 26 | projects/GLIP/configs/$exp3.py 27 | 28 | # Step4: test 29 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/test.py \ 30 | projects/GLIP/configs/$exp3.py \ 31 | work_dirs/$exp3/iter_10000.pth \ 32 | --work-dir work_dirs/$exp3/dior_test 33 | -------------------------------------------------------------------------------- /projects/GroundingDINO/groundingdino/__init__.py: -------------------------------------------------------------------------------- 1 | from .grounding_dino import RotatedGroundingDINO 2 | from .grounding_dino_head import RotatedGroundingDINOHead 3 | 4 | __all__ = [ 5 | 'RotatedGroundingDINO', 'RotatedGroundingDINOHead' 6 | ] -------------------------------------------------------------------------------- /projects/GroundingDINO/run.sh: -------------------------------------------------------------------------------- 1 | DEVICES_ID=1 2 | 3 | # Step1: train base-detector 4 | exp1="grounding_dino_swin-t_visdrone_base-set_adamw" 5 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/train.py \ 6 | projects/GroundingDINO/configs/$exp1.py 7 | 8 | # Step2.1: pseudo-labeling 9 | exp2="grounding_dino_swin-t_visdrone_base-set_adamw_nwpu45_pseudo_labeling" 10 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/test.py \ 11 | projects/GroundingDINO/configs/$exp2.py \ 12 | work_dirs/$exp1/iter_20000.pth 13 | 14 | # Step2.2: merge predictions 15 | python projects/GroundingDINO/tools/merge_ovdg_preds.py \ 16 | --ann_path data/NWPU-RESISC45/annotations/nwpu45_unlabeled_2.json \ 17 | --pred_path work_dirs/$exp2/nwpu45_pseudo_labeling_2.bbox.json \ 18 | --save_path work_dirs/$exp2/nwpu45_unlabeled_with_gdino_pseudos_swin-t_adamw_top1.json \ 19 | --topk 1 20 | 21 | cp work_dirs/$exp2/nwpu45_unlabeled_with_gdino_pseudos_swin-t_adamw_top1.json data/NWPU-RESISC45/annotations/nwpu45_unlabeled_with_gdino_pseudos_swin-t_adamw_top1.json 22 | 23 | # Step3: self-training 24 | exp3="grounding_dino_swin-t_visdrone_base-set_adamw_nwpu45" 25 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/train.py \ 26 | projects/GroundingDINO/configs/$exp3.py \ 27 | --work-dir work_dirs/$exp3 28 | 29 | # Step4: test 30 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/test.py \ 31 | projects/GroundingDINO/configs/$exp3.py \ 32 | work_dirs/$exp3/iter_10000.pth \ 33 | --work-dir work_dirs/$exp3/dior_test 34 | -------------------------------------------------------------------------------- /projects/LSKNet/lsknet/__init__.py: -------------------------------------------------------------------------------- 1 | from .lsknet import LSKNet 2 | 3 | __all__ = ['LSKNet'] 4 | -------------------------------------------------------------------------------- /projects/LabelStudio/backend_template/readme.md: -------------------------------------------------------------------------------- 1 | # Semi-automatic Object Detection Annotation with MMRotate and Label-Studio 2 | 3 | Please read the [Docs](../../docs/en/user_guides/label_studio.md) for more details. 4 | -------------------------------------------------------------------------------- /projects/RR360/configs360/.gitkeep: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/projects/RR360/configs360/.gitkeep -------------------------------------------------------------------------------- /projects/RR360/configs360/_base_/default_runtime.py: -------------------------------------------------------------------------------- 1 | default_scope = 'mmrotate' 2 | 3 | default_hooks = dict( 4 | timer=dict(type='IterTimerHook'), 5 | logger=dict(type='LoggerHook', interval=5), 6 | param_scheduler=dict(type='ParamSchedulerHook'), 7 | # checkpoint=dict(type='CheckpointHook', interval=1), 8 | checkpoint=dict( 9 | type='CheckpointHook', 10 | interval=4, 11 | save_best=['dota/AP50'], 12 | rule='greater', 13 | max_keep_ckpts=1), 14 | sampler_seed=dict(type='DistSamplerSeedHook'), 15 | visualization=dict(type='mmdet.DetVisualizationHook')) 16 | 17 | vis_backends = [ 18 | dict(type='LocalVisBackend'), 19 | # dict(type='WandbVisBackend', 20 | # init_kwargs=dict(project='trbox')) 21 | ] 22 | 23 | env_cfg = dict( 24 | cudnn_benchmark=False, 25 | mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0), 26 | dist_cfg=dict(backend='nccl'), 27 | ) 28 | 29 | visualizer = dict( 30 | type='RR360LocalVisualizer', vis_backends=vis_backends, name='visualizer') 31 | log_processor = dict(type='LogProcessor', window_size=50, by_epoch=True) 32 | 33 | log_level = 'INFO' 34 | load_from = None 35 | resume = False 36 | 37 | custom_imports = dict( 38 | imports=[ 39 | # 'mmcls.models', 40 | 'projects.RR360.visualization', 41 | 'projects.RR360.structures', 42 | 'projects.RR360.datasets.transforms', 43 | 'projects.RR360.evaluation', 44 | # 'projects.RR360.models' 45 | ], 46 | allow_failed_imports=False) 47 | -------------------------------------------------------------------------------- /projects/RR360/configs360/_base_/schedules/schedule_1x.py: -------------------------------------------------------------------------------- 1 | # training schedule for 1x 2 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=12, val_interval=1) 3 | val_cfg = dict(type='ValLoop') 4 | test_cfg = dict(type='TestLoop') 5 | 6 | # learning rate 7 | param_scheduler = [ 8 | dict( 9 | type='LinearLR', 10 | start_factor=1.0 / 3, 11 | by_epoch=False, 12 | begin=0, 13 | end=500), 14 | dict( 15 | type='MultiStepLR', 16 | begin=0, 17 | end=12, 18 | by_epoch=True, 19 | milestones=[8, 11], 20 | gamma=0.1) 21 | ] 22 | 23 | # optimizer 24 | optim_wrapper = dict( 25 | type='OptimWrapper', 26 | optimizer=dict(type='SGD', lr=0.0025, momentum=0.9, weight_decay=0.0001), 27 | clip_grad=dict(max_norm=35, norm_type=2)) 28 | -------------------------------------------------------------------------------- /projects/RR360/configs360/_base_/schedules/schedule_3x.py: -------------------------------------------------------------------------------- 1 | # training schedule for 1x 2 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=36, val_interval=1) 3 | val_cfg = dict(type='ValLoop') 4 | test_cfg = dict(type='TestLoop') 5 | 6 | # learning rate 7 | param_scheduler = [ 8 | dict( 9 | type='LinearLR', 10 | start_factor=1.0 / 3, 11 | by_epoch=False, 12 | begin=0, 13 | end=500), 14 | dict( 15 | type='MultiStepLR', 16 | begin=0, 17 | end=36, 18 | by_epoch=True, 19 | milestones=[24, 33], 20 | gamma=0.1) 21 | ] 22 | 23 | # optimizer 24 | optim_wrapper = dict( 25 | type='OptimWrapper', 26 | optimizer=dict(type='SGD', lr=0.0025, momentum=0.9, weight_decay=0.0001), 27 | clip_grad=dict(max_norm=35, norm_type=2)) 28 | -------------------------------------------------------------------------------- /projects/RR360/configs360/_base_/schedules/schedule_6x.py: -------------------------------------------------------------------------------- 1 | # training schedule for 1x 2 | train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=72, val_interval=6) 3 | val_cfg = dict(type='ValLoop') 4 | test_cfg = dict(type='TestLoop') 5 | 6 | # learning rate 7 | param_scheduler = [ 8 | dict( 9 | type='LinearLR', 10 | start_factor=1.0 / 3, 11 | by_epoch=False, 12 | begin=0, 13 | end=500), 14 | dict( 15 | type='MultiStepLR', 16 | begin=0, 17 | end=72, 18 | by_epoch=True, 19 | milestones=[48, 66], 20 | gamma=0.1) 21 | ] 22 | 23 | # optimizer 24 | optim_wrapper = dict( 25 | type='OptimWrapper', 26 | optimizer=dict( 27 | type='SGD', lr=0.0025 / 2, momentum=0.9, weight_decay=0.0001), 28 | clip_grad=dict(max_norm=35, norm_type=2)) 29 | -------------------------------------------------------------------------------- /projects/RR360/configs360/readme.md: -------------------------------------------------------------------------------- 1 | # configs for 360 detection 2 | -------------------------------------------------------------------------------- /projects/RR360/configs360/rotated_rtmdet_x3/_base_/default_runtime.py: -------------------------------------------------------------------------------- 1 | default_scope = 'mmrotate' 2 | 3 | default_hooks = dict( 4 | timer=dict(type='IterTimerHook'), 5 | logger=dict(type='LoggerHook', interval=5), 6 | param_scheduler=dict(type='ParamSchedulerHook'), 7 | checkpoint=dict(type='CheckpointHook', interval=12, max_keep_ckpts=3), 8 | sampler_seed=dict(type='DistSamplerSeedHook'), 9 | visualization=dict(type='mmdet.DetVisualizationHook')) 10 | 11 | env_cfg = dict( 12 | cudnn_benchmark=False, 13 | mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0), 14 | dist_cfg=dict(backend='nccl'), 15 | ) 16 | 17 | vis_backends = [ 18 | dict(type='LocalVisBackend'), 19 | # dict(type='WandbVisBackend', init_kwargs=dict(project='trbox')) 20 | ] 21 | 22 | visualizer = dict( 23 | type='RR360LocalVisualizer', vis_backends=vis_backends, name='visualizer') 24 | log_processor = dict(type='LogProcessor', window_size=50, by_epoch=True) 25 | 26 | log_level = 'INFO' 27 | load_from = None 28 | resume = False 29 | 30 | custom_hooks = [ 31 | dict(type='mmdet.NumClassCheckHook'), 32 | dict( 33 | type='EMAHook', 34 | ema_type='mmdet.ExpMomentumEMA', 35 | momentum=0.0002, 36 | update_buffers=True, 37 | priority=49) 38 | ] 39 | 40 | custom_imports = dict( 41 | imports=[ 42 | # 'mmcls.models', 43 | 'projects.RR360.visualization', 44 | 'projects.RR360.structures', 45 | 'projects.RR360.datasets.transforms', 46 | 'projects.RR360.evaluation', 47 | # 'projects.RR360.models', 48 | ], 49 | allow_failed_imports=False) 50 | -------------------------------------------------------------------------------- /projects/RR360/configs360/rotated_rtmdet_x3/_base_/schedule_3x.py: -------------------------------------------------------------------------------- 1 | max_epochs = 3 * 12 2 | base_lr = 0.004 / 16 3 | interval = 6 4 | 5 | train_cfg = dict( 6 | type='EpochBasedTrainLoop', max_epochs=max_epochs, val_interval=interval) 7 | val_cfg = dict(type='ValLoop') 8 | test_cfg = dict(type='TestLoop') 9 | 10 | # learning rate 11 | param_scheduler = [ 12 | dict( 13 | type='LinearLR', 14 | start_factor=1.0e-5, 15 | by_epoch=False, 16 | begin=0, 17 | end=1000), 18 | dict( 19 | type='CosineAnnealingLR', 20 | eta_min=base_lr * 0.05, 21 | begin=max_epochs // 2, 22 | end=max_epochs, 23 | T_max=max_epochs // 2, 24 | by_epoch=True, 25 | convert_to_iter_based=True), 26 | ] 27 | 28 | # optimizer 29 | optim_wrapper = dict( 30 | type='OptimWrapper', 31 | optimizer=dict(type='AdamW', lr=base_lr, weight_decay=0.05), 32 | paramwise_cfg=dict( 33 | norm_decay_mult=0, bias_decay_mult=0, bypass_duplicate=True)) 34 | -------------------------------------------------------------------------------- /projects/RR360/configs360/rotated_rtmdet_x3_r/_base_/default_runtime.py: -------------------------------------------------------------------------------- 1 | default_scope = 'mmrotate' 2 | 3 | default_hooks = dict( 4 | timer=dict(type='IterTimerHook'), 5 | logger=dict(type='LoggerHook', interval=5), 6 | param_scheduler=dict(type='ParamSchedulerHook'), 7 | checkpoint=dict(type='CheckpointHook', interval=12, max_keep_ckpts=3), 8 | sampler_seed=dict(type='DistSamplerSeedHook'), 9 | visualization=dict(type='mmdet.DetVisualizationHook')) 10 | 11 | env_cfg = dict( 12 | cudnn_benchmark=False, 13 | mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0), 14 | dist_cfg=dict(backend='nccl'), 15 | ) 16 | 17 | vis_backends = [ 18 | dict(type='LocalVisBackend'), 19 | # dict(type='WandbVisBackend', init_kwargs=dict(project='trbox')) 20 | ] 21 | 22 | visualizer = dict( 23 | type='RR360LocalVisualizer', vis_backends=vis_backends, name='visualizer') 24 | log_processor = dict(type='LogProcessor', window_size=50, by_epoch=True) 25 | 26 | log_level = 'INFO' 27 | load_from = None 28 | resume = False 29 | 30 | custom_hooks = [ 31 | dict(type='mmdet.NumClassCheckHook'), 32 | dict( 33 | type='EMAHook', 34 | ema_type='mmdet.ExpMomentumEMA', 35 | momentum=0.0002, 36 | update_buffers=True, 37 | priority=49) 38 | ] 39 | 40 | custom_imports = dict( 41 | imports=[ 42 | # 'mmcls.models', 43 | 'projects.RR360.visualization', 44 | 'projects.RR360.structures', 45 | 'projects.RR360.datasets.transforms', 46 | 'projects.RR360.evaluation', 47 | # 'projects.RR360.models', 48 | ], 49 | allow_failed_imports=False) 50 | -------------------------------------------------------------------------------- /projects/RR360/configs360/rotated_rtmdet_x3_r/_base_/schedule_3x.py: -------------------------------------------------------------------------------- 1 | max_epochs = 3 * 12 2 | base_lr = 0.004 / 16 3 | interval = 6 4 | 5 | train_cfg = dict( 6 | type='EpochBasedTrainLoop', max_epochs=max_epochs, val_interval=interval) 7 | val_cfg = dict(type='ValLoop') 8 | test_cfg = dict(type='TestLoop') 9 | 10 | # learning rate 11 | param_scheduler = [ 12 | dict( 13 | type='LinearLR', 14 | start_factor=1.0e-5, 15 | by_epoch=False, 16 | begin=0, 17 | end=1000), 18 | dict( 19 | type='CosineAnnealingLR', 20 | eta_min=base_lr * 0.05, 21 | begin=max_epochs // 2, 22 | end=max_epochs, 23 | T_max=max_epochs // 2, 24 | by_epoch=True, 25 | convert_to_iter_based=True), 26 | ] 27 | 28 | # optimizer 29 | optim_wrapper = dict( 30 | type='OptimWrapper', 31 | optimizer=dict(type='AdamW', lr=base_lr, weight_decay=0.05), 32 | paramwise_cfg=dict( 33 | norm_decay_mult=0, bias_decay_mult=0, bypass_duplicate=True)) 34 | -------------------------------------------------------------------------------- /projects/RR360/datasets/__init__.py: -------------------------------------------------------------------------------- 1 | from .transforms import * # noqa: F401, F403 2 | -------------------------------------------------------------------------------- /projects/RR360/datasets/transforms/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .transforms import RotateAutoBound 3 | 4 | __all__ = ['RotateAutoBound'] 5 | -------------------------------------------------------------------------------- /projects/RR360/evaluation/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .functional import * # noqa: F401,F403 3 | from .metrics import * # noqa: F401,F403 4 | -------------------------------------------------------------------------------- /projects/RR360/evaluation/functional/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .mean_ap import eval_rbbox_head_map 3 | 4 | __all__ = ['eval_rbbox_head_map'] 5 | -------------------------------------------------------------------------------- /projects/RR360/evaluation/metrics/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .dota_r360_metric import DOTAR360Metric 3 | 4 | __all__ = ['DOTAR360Metric'] 5 | -------------------------------------------------------------------------------- /projects/RR360/structures/__init__.py: -------------------------------------------------------------------------------- 1 | from .bbox import * # noqa: F401,F403 2 | -------------------------------------------------------------------------------- /projects/RR360/structures/bbox/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | 3 | # from .rotated_boxes import RotatedBoxes, rbox2qbox 4 | from .rotated_boxes import RotatedBoxes 5 | 6 | __all__ = ['RotatedBoxes'] 7 | # __all__ = [ 8 | # 'RotatedBoxes', 'rbox2qbox' 9 | # ] 10 | -------------------------------------------------------------------------------- /projects/RR360/visualization/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from .local_visualizer import RR360LocalVisualizer 3 | 4 | __all__ = ['RR360LocalVisualizer'] 5 | -------------------------------------------------------------------------------- /projects/ViLD/resources/vild_framework.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/projects/ViLD/resources/vild_framework.png -------------------------------------------------------------------------------- /projects/ViLD/run.sh: -------------------------------------------------------------------------------- 1 | DEVICES_ID=0 2 | 3 | # Step1: train the base model 4 | exp1="oriented-rcnn_r50-fpn_20k_visdronezsd_base-set" 5 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/train.py \ 6 | projects/ViLD/configs/$exp1.py 7 | 8 | ## Step2: merge weights 9 | python projects/CastDetv2/tools/merge_weights.py \ 10 | --clip_path checkpoints/RemoteCLIP-RN50.pt \ 11 | --base_path work_dirs/$exp1/iter_20000.pth \ 12 | --save_path work_dirs/$exp1/merged_vild_init_iter20k.pth 13 | --target_model vild 14 | 15 | 16 | # Step3: prepare pseudo labels 17 | exp2="vild_oriented-rcnn_r50_fpn_visdronezsd_step1_prepare" 18 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/test.py \ 19 | projects/ViLD/configs/$exp2.py \ 20 | work_dirs/$exp1/merged_vild_init_iter20k.pth 21 | 22 | # Step4: self-training 23 | exp3="vild_oriented-rcnn_r50_fpn_visdronezsd_step2_finetune" 24 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/train.py \ 25 | projects/ViLD/configs/$exp3.py 26 | 27 | # Step5: test 28 | CUDA_VISIBLE_DEVICES=$DEVICES_ID python tools/test.py \ 29 | projects/ViLD/configs/$exp3.py \ 30 | work_dirs/$exp3/iter_10000.pth \ 31 | --work-dir work_dirs/$exp3/dior_test 32 | 33 | -------------------------------------------------------------------------------- /projects/ViLD/vild/__init__.py: -------------------------------------------------------------------------------- 1 | from .rotated_vild import RotatedViLD 2 | from .ovd_bbox_head import Shared2FCBBoxHeadZSD, Projection2 3 | from .modified_resnet import ModifiedResNet2 4 | 5 | __all__ = [ 6 | 'RotatedViLD', 'Shared2FCBBoxHeadZSD', 'Projection2' 7 | 'ModifiedResNet2' 8 | ] -------------------------------------------------------------------------------- /projects/example_project/configs/r3det-oc_dummy-resnet_fpn_1x_dota.py: -------------------------------------------------------------------------------- 1 | _base_ = ['../../../configs/r3det/r3det-oc_r50_fpn_1x_dota.py'] 2 | 3 | custom_imports = dict(imports=['projects.example_project.dummy']) 4 | 5 | _base_.model.backbone.type = 'DummyResNet' 6 | -------------------------------------------------------------------------------- /projects/example_project/dummy/__init__.py: -------------------------------------------------------------------------------- 1 | from .dummy_resnet import DummyResNet 2 | 3 | __all__ = ['DummyResNet'] 4 | -------------------------------------------------------------------------------- /projects/example_project/dummy/dummy_resnet.py: -------------------------------------------------------------------------------- 1 | from mmdet.models.backbones import ResNet 2 | 3 | from mmrotate.registry import MODELS 4 | 5 | 6 | @MODELS.register_module() 7 | class DummyResNet(ResNet): 8 | """Implements a dummy ResNet wrapper for demonstration purpose. 9 | Args: 10 | **kwargs: All the arguments are passed to the parent class. 11 | """ 12 | 13 | def __init__(self, **kwargs) -> None: 14 | print('Hello world!') 15 | super().__init__(**kwargs) 16 | -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | -r requirements/build.txt 2 | -r requirements/optional.txt 3 | -r requirements/runtime.txt 4 | -r requirements/tests.txt 5 | -------------------------------------------------------------------------------- /requirements/build.txt: -------------------------------------------------------------------------------- 1 | # These must be installed before building mmrotate 2 | cython 3 | numpy 4 | -------------------------------------------------------------------------------- /requirements/docs.txt: -------------------------------------------------------------------------------- 1 | docutils==0.16.0 2 | myst-parser 3 | -e git+https://github.com/open-mmlab/pytorch_sphinx_theme.git#egg=pytorch_sphinx_theme 4 | sphinx==4.0.2 5 | sphinx-copybutton 6 | sphinx_markdown_tables 7 | sphinx_rtd_theme==0.5.2 8 | -------------------------------------------------------------------------------- /requirements/mminstall.txt: -------------------------------------------------------------------------------- 1 | mmcv>=2.0.0rc2,<2.1.0 2 | mmdet>=3.0.0rc2,<3.2.0 3 | mmengine>= 0.1.0 4 | -------------------------------------------------------------------------------- /requirements/multimodal.txt: -------------------------------------------------------------------------------- 1 | fairscale 2 | jsonlines 3 | nltk 4 | pycocoevalcap 5 | transformers 6 | -------------------------------------------------------------------------------- /requirements/optional.txt: -------------------------------------------------------------------------------- 1 | imagecorruptions 2 | scikit-learn 3 | scipy 4 | -------------------------------------------------------------------------------- /requirements/readthedocs.txt: -------------------------------------------------------------------------------- 1 | e2cnn 2 | mmcv>=2.0.0rc2 3 | mmdet>=3.0.0rc2 4 | mmengine>=0.1.0 5 | torch 6 | torchvision 7 | -------------------------------------------------------------------------------- /requirements/runtime.txt: -------------------------------------------------------------------------------- 1 | matplotlib 2 | numpy 3 | pycocotools 4 | six 5 | terminaltables 6 | torch 7 | -------------------------------------------------------------------------------- /requirements/tests.txt: -------------------------------------------------------------------------------- 1 | asynctest 2 | codecov 3 | coverage 4 | cython 5 | -e git+https://github.com/QUVA-Lab/e2cnn.git#egg=e2cnn 6 | flake8 7 | interrogate 8 | isort==4.3.21 9 | # Note: used for kwarray.group_items, this may be ported to mmcv in the future. 10 | kwarray 11 | matplotlib 12 | parameterized 13 | pytest 14 | scikit-learn 15 | ubelt 16 | wheel 17 | xdoctest>=0.10.0 18 | yapf 19 | -------------------------------------------------------------------------------- /resources/mmrotate-logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/resources/mmrotate-logo.png -------------------------------------------------------------------------------- /resources/qq_group_qrcode.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/resources/qq_group_qrcode.jpg -------------------------------------------------------------------------------- /resources/zhihu_qrcode.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/resources/zhihu_qrcode.jpg -------------------------------------------------------------------------------- /setup.cfg: -------------------------------------------------------------------------------- 1 | [isort] 2 | line_length = 79 3 | multi_line_output = 0 4 | known_standard_library = setuptools 5 | known_first_party = mmrotate 6 | known_third_party = PIL,boto3,botocore,cv2,e2cnn,label_studio_ml,label_studio_tools,matplotlib,mmcv,mmdet,mmengine,numpy,parameterized,pycocotools,pytest,pytorch_sphinx_theme,terminaltables,torch,torchvision,ts,yaml 7 | no_lines_before = STDLIB,LOCALFOLDER 8 | default_section = THIRDPARTY 9 | 10 | [yapf] 11 | BASED_ON_STYLE = pep8 12 | BLANK_LINE_BEFORE_NESTED_CLASS_OR_DEF = true 13 | SPLIT_BEFORE_EXPRESSION_AFTER_OPENING_PAREN = true 14 | 15 | [codespell] 16 | skip = *.ipynb 17 | quiet-level = 3 18 | ignore-words-list = DOTA,dota,alse, warmup 19 | -------------------------------------------------------------------------------- /tests/data/dior/Annotations/Oriented Bounding Boxes/00001.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | /JPEGImages-trainval.zip/JPEGImages-trainval/JPEGImages-trainval 4 | 00001.jpg 5 | /JPEGImages-trainval.zip/JPEGImages-trainval/JPEGImages-trainval 6 | 7 | Unknown 8 | 9 | 10 | 800 11 | 800 12 | 3 13 | 14 | 0 15 | 16 | robndbox 17 | golffield 18 | Unspecified 19 | 0 20 | 0 21 | 22 | 133 23 | 237 24 | 684 25 | 237 26 | 684 27 | 672 28 | 133 29 | 672 30 | 31 | 0 32 | 33 | 34 | -------------------------------------------------------------------------------- /tests/data/dior/JPEGImage/00001.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/tests/data/dior/JPEGImage/00001.jpg -------------------------------------------------------------------------------- /tests/data/dior/demo.txt: -------------------------------------------------------------------------------- 1 | 00001 2 | -------------------------------------------------------------------------------- /tests/data/dota/images/P2805__1024__0___0.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/VisionXLab/CastDet/32e67b53bf739208a69615c91b2789f1be109536/tests/data/dota/images/P2805__1024__0___0.png -------------------------------------------------------------------------------- /tests/data/dota/labelTxt/P2805__1024__0___0.txt: -------------------------------------------------------------------------------- 1 | 359.0 663.0 369.0 497.0 543.0 509.0 531.0 677.0 plane 0 2 | 540.0 884.0 363.0 862.0 392.0 674.0 570.0 695.0 plane 0 3 | 788.0 844.0 734.0 701.0 916.0 631.0 970.0 762.0 plane 0 4 | 720.0 726.0 668.0 583.0 852.0 494.0 913.0 636.0 plane 0 5 | -------------------------------------------------------------------------------- /tests/data/hrsc/FullDataSet/Annotations/100000006.xml: -------------------------------------------------------------------------------- 1 | 2 | 100000006 3 | 100000001 4 | 100000001 5 | 100000006 6 | 100000006 7 | bmp 8 | 1900-01-01 9 | sealand 10 | 11 | 12 | 69.040297,33.070036 13 | 1172 14 | 816 15 | 3 16 | 1.07 17 | 18 18 | 100 19 | 20 | 21 | 0 22 | 0 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 000d 34 | 1 35 | 36 | 37 | 100000006 38 | 100000013 39 | 100000006 40 | 0 41 | 0 42 | 119 43 | 75 44 | 587 45 | 789 46 | 341.2143 47 | 443.3325 48 | 778.4297 49 | 178.2595 50 | -1.122944 51 | 0 52 | 53 | 54 | 143 55 | 776 56 | 57 | 58 | 59 | -------------------------------------------------------------------------------- /tests/data/hrsc/demo.txt: -------------------------------------------------------------------------------- 1 | 100000006 2 | -------------------------------------------------------------------------------- /tests/test_apis/test_inference.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | import os.path as osp 3 | from unittest import TestCase 4 | 5 | import numpy as np 6 | import pytest 7 | import torch 8 | from mmdet.apis import init_detector 9 | from mmdet.structures import DetDataSample 10 | from parameterized import parameterized 11 | 12 | from mmrotate.apis import inference_detector_by_patches 13 | from mmrotate.utils import register_all_modules 14 | 15 | 16 | class TestInferenceDetectorByPatches(TestCase): 17 | 18 | def setUp(self): 19 | register_all_modules() 20 | 21 | @parameterized.expand([ 22 | ('rotated_retinanet/rotated-retinanet-rbox-le90_r50_fpn_1x_dota.py', 23 | ('cuda', )), 24 | ]) 25 | def test_inference_detector_by_patches(self, config, devices): 26 | assert all([device in ['cpu', 'cuda'] for device in devices]) 27 | 28 | project_dir = osp.abspath(osp.dirname(osp.dirname(__file__))) 29 | project_dir = osp.join(project_dir, '..') 30 | 31 | config_file = osp.join(project_dir, 'configs', config) 32 | 33 | # test init_detector with config_file: str and cfg_options 34 | rng = np.random.RandomState(0) 35 | img = rng.randint(0, 255, (125, 125, 3), dtype=np.uint8) 36 | 37 | for device in devices: 38 | if device == 'cuda' and not torch.cuda.is_available(): 39 | pytest.skip('test requires GPU and torch+cuda') 40 | 41 | model = init_detector(config_file, device=device) 42 | nms_cfg = dict(type='nms_rotated', iou_threshold=0.1) 43 | result = inference_detector_by_patches(model, img, [75], [50], 44 | [1.0], nms_cfg) 45 | assert isinstance(result, DetDataSample) 46 | -------------------------------------------------------------------------------- /tests/test_datasets/test_dior.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | import unittest 3 | 4 | from mmrotate.datasets import DIORDataset 5 | 6 | 7 | class TestDIORDataset(unittest.TestCase): 8 | 9 | def test_dior(self): 10 | dataset = DIORDataset( 11 | data_root='tests/data/dior/', 12 | ann_file='demo.txt', 13 | data_prefix=dict(img_path='JPEGImages/'), 14 | filter_cfg=dict( 15 | filter_empty_gt=True, min_size=32, bbox_min_size=4), 16 | pipeline=[]) 17 | dataset.full_init() 18 | self.assertEqual(len(dataset), 1) 19 | 20 | data_list = dataset.load_data_list() 21 | self.assertEqual(len(data_list), 1) 22 | self.assertEqual(data_list[0]['img_id'], '00001') 23 | self.assertEqual(data_list[0]['img_path'].replace('\\', '/'), 24 | 'tests/data/dior/JPEGImages/00001.jpg') 25 | self.assertEqual( 26 | data_list[0]['xml_path'].replace('\\', '/'), 27 | 'tests/data/dior/Annotations/Oriented Bounding Boxes/00001.xml') 28 | self.assertEqual(len(data_list[0]['instances']), 1) 29 | self.assertEqual(dataset.get_cat_ids(0), [9]) 30 | self.assertEqual(len(dataset._metainfo['classes']), 20) 31 | -------------------------------------------------------------------------------- /tests/test_models/test_roi_heads/test_bbox_head/test_gv_bbox_head.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | import unittest 3 | from unittest import TestCase 4 | 5 | import torch 6 | from parameterized import parameterized 7 | 8 | from mmrotate.models.roi_heads.bbox_heads import GVBBoxHead 9 | from mmrotate.utils import register_all_modules 10 | 11 | 12 | class TestGVBBoxHead(TestCase): 13 | 14 | def setUp(self): 15 | register_all_modules() 16 | 17 | @parameterized.expand(['cpu', 'cuda']) 18 | def test_forward_loss(self, device): 19 | if device == 'cuda': 20 | if not torch.cuda.is_available(): 21 | return unittest.skip('test requires GPU and torch+cuda') 22 | 23 | bbox_head = GVBBoxHead( 24 | in_channels=1, 25 | fc_out_channels=1, 26 | roi_feat_size=7, 27 | num_classes=4, 28 | ratio_thr=0.8, 29 | bbox_coder=dict( 30 | type='DeltaXYWHQBBoxCoder', 31 | target_means=(.0, .0, .0, .0), 32 | target_stds=(0.1, 0.1, 0.2, 0.2)), 33 | fix_coder=dict(type='GVFixCoder'), 34 | ratio_coder=dict(type='GVRatioCoder'), 35 | predict_box_type='rbox', 36 | reg_class_agnostic=True, 37 | loss_cls=dict( 38 | type='mmdet.CrossEntropyLoss', 39 | use_sigmoid=False, 40 | loss_weight=1.0), 41 | loss_bbox=dict( 42 | type='mmdet.SmoothL1Loss', beta=1.0, loss_weight=1.0), 43 | loss_fix=dict( 44 | type='mmdet.SmoothL1Loss', beta=1.0 / 3.0, loss_weight=1.0), 45 | loss_ratio=dict( 46 | type='mmdet.SmoothL1Loss', beta=1.0 / 3.0, loss_weight=16.0)) 47 | bbox_head = bbox_head.to(device=device) 48 | 49 | num_samples = 4 50 | feats = torch.rand((num_samples, 1, 7, 7)).to(device) 51 | bbox_head(x=feats) 52 | -------------------------------------------------------------------------------- /tests/test_models/test_task_modules/test_assigners/test_convex_assigner.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | import unittest 3 | 4 | import torch 5 | from mmengine.structures import InstanceData 6 | from mmengine.testing import assert_allclose 7 | 8 | from mmrotate.models.task_modules.assigners import ConvexAssigner 9 | 10 | 11 | class TestConvexAssigner(unittest.TestCase): 12 | 13 | def test_convex_assigner(self): 14 | if not torch.cuda.is_available(): 15 | return unittest.skip('test requires GPU and torch+cuda') 16 | 17 | assigner = ConvexAssigner(scale=4, pos_num=1) 18 | priors = torch.FloatTensor([ 19 | [0, 0, 2, 2, 5, 5, 10, 10, 10, 5, 10, 0, 8, 0, 5, 0, 2, 0], 20 | [10, 0, 12, 2, 15, 5, 20, 10, 20, 5, 20, 0, 18, 0, 15, 0, 12, 0], 21 | [ 22 | 10, 10, 12, 12, 15, 15, 20, 20, 20, 15, 20, 10, 18, 10, 15, 10, 23 | 12, 10 24 | ], 25 | [ 26 | 12, 10, 14, 12, 17, 15, 22, 20, 22, 15, 22, 10, 20, 10, 17, 10, 27 | 14, 10 28 | ], 29 | ]).cuda() 30 | gt_bboxes = torch.FloatTensor([ 31 | [5, 5, 10, 5, 10, 10, 5, 10], 32 | [5, 15, 15, 15, 15, 20, 5, 20], 33 | ]).cuda() 34 | gt_labels = torch.LongTensor([2, 3]).cuda() 35 | 36 | pred_instances = InstanceData(priors=priors) 37 | gt_instances = InstanceData(bboxes=gt_bboxes, labels=gt_labels) 38 | 39 | assign_result = assigner.assign(pred_instances, gt_instances) 40 | self.assertEqual(len(assign_result.gt_inds), 4) 41 | self.assertEqual(len(assign_result.labels), 4) 42 | 43 | expected_gt_inds = torch.LongTensor([1, 0, 0, 0]).cuda() 44 | assert_allclose(assign_result.gt_inds, expected_gt_inds) 45 | -------------------------------------------------------------------------------- /tests/test_models/test_task_modules/test_assigners/test_max_convex_iou_assigner.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | import unittest 3 | 4 | import torch 5 | from mmengine.structures import InstanceData 6 | from parameterized import parameterized 7 | 8 | from mmrotate.models.task_modules.assigners import MaxConvexIoUAssigner 9 | 10 | 11 | class TestMaxConvexIoUAssigner(unittest.TestCase): 12 | 13 | @parameterized.expand([(0.5, ), ((0, 0.5), )]) 14 | def test_max_convex_iou_assigner(self, neg_iou_thr): 15 | 16 | if not torch.cuda.is_available(): 17 | return unittest.skip('test requires GPU and torch+cuda') 18 | 19 | assigner = MaxConvexIoUAssigner( 20 | pos_iou_thr=0.5, 21 | neg_iou_thr=neg_iou_thr, 22 | ) 23 | priors = torch.FloatTensor([ 24 | [0, 0, 2, 2, 5, 5, 10, 10, 10, 5, 10, 0, 8, 0, 5, 0, 2, 0], 25 | [10, 0, 12, 2, 15, 5, 20, 10, 20, 5, 20, 0, 18, 0, 15, 0, 12, 0], 26 | [ 27 | 10, 10, 12, 12, 15, 15, 20, 20, 20, 15, 20, 10, 18, 10, 15, 10, 28 | 12, 10 29 | ], 30 | [ 31 | 12, 10, 14, 12, 17, 15, 22, 20, 22, 15, 22, 10, 20, 10, 17, 10, 32 | 14, 10 33 | ], 34 | ]).cuda() 35 | gt_bboxes = torch.FloatTensor([ 36 | [5, 5, 5, 4, 0.1], 37 | [5, 15, 5, 5, 0.0], 38 | ]).cuda() 39 | gt_labels = torch.LongTensor([2, 3]).cuda() 40 | 41 | pred_instances = InstanceData(priors=priors) 42 | gt_instances = InstanceData(bboxes=gt_bboxes, labels=gt_labels) 43 | 44 | assign_result = assigner.assign(pred_instances, gt_instances) 45 | self.assertEqual(len(assign_result.gt_inds), 4) 46 | self.assertEqual(len(assign_result.labels), 4) 47 | -------------------------------------------------------------------------------- /tests/test_models/test_task_modules/test_coder/test_angle_coder.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from unittest import TestCase 3 | 4 | import torch 5 | from mmengine.testing import assert_allclose 6 | 7 | from mmrotate.models.task_modules.coders import CSLCoder 8 | 9 | 10 | class TestCSLCoder(TestCase): 11 | 12 | def test_encode(self): 13 | coder = CSLCoder(angle_version='oc', omega=10) 14 | angle_preds = torch.Tensor([[0.]]) 15 | expected_encode_angles = torch.Tensor( 16 | [[0., 0., 0., 0., 0., 0., 0., 0., 0.]]) 17 | out = coder.encode(angle_preds) 18 | assert_allclose(expected_encode_angles, out) 19 | 20 | def test_decode(self): 21 | coder = CSLCoder(angle_version='oc', omega=10) 22 | encode_angles = torch.Tensor([[0., 0., 0., 0., 0., 0., 0., 0., 0.]]) 23 | out = coder.decode(encode_angles) 24 | assert_allclose([1], out.shape) 25 | -------------------------------------------------------------------------------- /tests/test_models/test_task_modules/test_coder/test_delta_xywh_hbbox_coder.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from unittest import TestCase 3 | 4 | import torch 5 | from mmdet.structures.bbox import HorizontalBoxes 6 | from mmengine.testing import assert_allclose 7 | 8 | from mmrotate.models.task_modules.coders import DeltaXYWHHBBoxCoder 9 | from mmrotate.structures.bbox import RotatedBoxes 10 | 11 | 12 | class TestDeltaBboxCoder(TestCase): 13 | 14 | def test_encode(self): 15 | coder = DeltaXYWHHBBoxCoder() 16 | 17 | proposals = torch.Tensor([[0., 0., 1., 1.], [0., 0., 2., 2.], 18 | [0., 0., 5., 5.], [5., 5., 10., 10.]]) 19 | gt = torch.Tensor([[0., 0., 1., 1., 0.], [0., 0., 1., 1., 0.1], 20 | [0., 0., 1., 1., 0.1], [5., 5., 5., 5., 0.3]]) 21 | 22 | expected_encode_bboxes = torch.Tensor( 23 | [[-0.5000, -0.5000, 0.0000, 0.0000], 24 | [-0.5000, -0.5000, -0.6025, -0.6025], 25 | [-0.5000, -0.5000, -1.5188, -1.5188], 26 | [-0.5000, -0.5000, 0.2238, 0.2238]]) 27 | 28 | out = coder.encode(HorizontalBoxes(proposals), RotatedBoxes(gt)) 29 | assert_allclose(expected_encode_bboxes, out) 30 | -------------------------------------------------------------------------------- /tests/test_models/test_task_modules/test_coder/test_delta_xywh_qbbox_coder.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from unittest import TestCase 3 | 4 | import torch 5 | from mmdet.structures.bbox import HorizontalBoxes 6 | from mmengine.testing import assert_allclose 7 | 8 | from mmrotate.models.task_modules.coders import DeltaXYWHQBBoxCoder 9 | from mmrotate.structures.bbox import QuadriBoxes 10 | 11 | 12 | class TestDeltaXYWHQBBoxCoder(TestCase): 13 | 14 | def test_encode(self): 15 | coder = DeltaXYWHQBBoxCoder() 16 | 17 | proposals = torch.Tensor([[0., 0., 1., 1.], [0., 0., 2., 2.], 18 | [0., 0., 5., 5.], [5., 5., 10., 10.]]) 19 | gt = torch.Tensor([[0., 0., 1., 0., 1., 1., 0., 1.], 20 | [0.1, 0., 1.1, 0., 1.1, 1., 0.1, 1.], 21 | [0., 0.1, 1., 0.1, 1., 1.1, 0., 1.1], 22 | [0.1, 0.1, 1.1, 0.1, 1.1, 1.1, 0.1, 1.1]]) 23 | 24 | expected_encode_bboxes = torch.Tensor( 25 | [[0.0000, 0.0000, 0.0000, 0.0000], 26 | [-0.2000, -0.2500, -0.6931, -0.6931], 27 | [-0.4000, -0.3800, -1.6094, -1.6094], 28 | [-1.3800, -1.3800, -1.6094, -1.6094]]) 29 | 30 | out = coder.encode(HorizontalBoxes(proposals), QuadriBoxes(gt)) 31 | assert_allclose(expected_encode_bboxes, out) 32 | -------------------------------------------------------------------------------- /tests/test_visualization/test_palette.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | from mmrotate.datasets import DOTADataset 3 | from mmrotate.visualization import get_palette 4 | 5 | 6 | def test_palette(): 7 | 8 | # test list 9 | palette = [(1, 0, 0), (0, 1, 0), (0, 0, 1)] 10 | palette_ = get_palette(palette, 3) 11 | for color, color_ in zip(palette, palette_): 12 | assert color == color_ 13 | 14 | # test tuple 15 | palette = get_palette((1, 2, 3), 3) 16 | assert len(palette) == 3 17 | for color in palette: 18 | assert color == (1, 2, 3) 19 | 20 | # test color str 21 | palette = get_palette('red', 3) 22 | assert len(palette) == 3 23 | for color in palette: 24 | assert color == (255, 0, 0) 25 | 26 | # test dataset str 27 | palette = get_palette('dota', len(DOTADataset.METAINFO['classes'])) 28 | assert len(palette) == len(DOTADataset.METAINFO['classes']) 29 | assert palette[0] == (165, 42, 42) 30 | 31 | # test random 32 | palette1 = get_palette('random', 3) 33 | palette2 = get_palette(None, 3) 34 | for color1, color2 in zip(palette1, palette2): 35 | assert isinstance(color1, tuple) 36 | assert isinstance(color2, tuple) 37 | assert color1 == color2 38 | -------------------------------------------------------------------------------- /tools/data/README.md: -------------------------------------------------------------------------------- 1 | # Data Preparation for Rotation Detection 2 | 3 | It is recommended to symlink the dataset root to `$MMROTATE/data`. 4 | If your folder structure is different, you may need to change the corresponding paths in config files. 5 | 6 | Datasets supported in MMRotate: 7 | 8 | - [DOTA Dataset](dota/README.md) \[ [Homepage](https://captain-whu.github.io/DOTA/) \] 9 | - [DIOR Dataset](dior/README.md) \[ [Homepage](https://gcheng-nwpu.github.io/#Datasets) \] 10 | - [SSDD Dataset](ssdd/README.md) 11 | - [HRSC Dataset](hrsc/README.md) 12 | - [HRSID Dataset](hrsid/README.md) 13 | - [SRSDD Dataset](srsdd/README.md) 14 | - [RSDD Dataset](rsdd/README.md) 15 | -------------------------------------------------------------------------------- /tools/data/dior/README.md: -------------------------------------------------------------------------------- 1 | # Preparing DIOR Dataset 2 | 3 | 4 | 5 | ```bibtex 6 | @article{LI2020296, 7 | title = {Object detection in optical remote sensing images: A survey and a new benchmark}, 8 | journal = {ISPRS Journal of Photogrammetry and Remote Sensing}, 9 | volume = {159}, 10 | pages = {296-307}, 11 | year = {2020}, 12 | issn = {0924-2716}, 13 | doi = {https://doi.org/10.1016/j.isprsjprs.2019.11.023}, 14 | url = {https://www.sciencedirect.com/science/article/pii/S0924271619302825}, 15 | author = {Ke Li and Gang Wan and Gong Cheng and Liqiu Meng and Junwei Han} 16 | ``` 17 | 18 | ## Download DIOR dataset 19 | 20 | The DIOR dataset can be downloaded from [here](https://gcheng-nwpu.github.io/#Datasets). 21 | 22 | The data structure is as follows: 23 | 24 | ```none 25 | mmrotate 26 | ├── mmrotate 27 | ├── tools 28 | ├── configs 29 | ├── data 30 | │ ├── DIOR 31 | │ │ ├── JPEGImages-trainval 32 | │ │ ├── JPEGImages-test 33 | │ │ ├── Annotations 34 | │ │ │ ├─ Oriented Bounding Boxes 35 | │ │ │ ├─ Horizontal Bounding Boxes 36 | │ │ ├── ImageSets 37 | │ │ │ ├─ Main 38 | │ │ │ │ ├─ train.txt 39 | │ │ │ │ ├─ val.txt 40 | │ │ │ │ ├─ test.txt 41 | ``` 42 | 43 | ## Change base config 44 | 45 | Please change `data_root` in `configs/_base_/datasets/dior.py` to `data/dior/`. 46 | -------------------------------------------------------------------------------- /tools/data/dota/split/split_configs/ms_test.json: -------------------------------------------------------------------------------- 1 | { 2 | "nproc": 10, 3 | "img_dirs": [ 4 | "data/DOTA/test/images/" 5 | ], 6 | "ann_dirs": null, 7 | "sizes": [ 8 | 1024 9 | ], 10 | "gaps": [ 11 | 500 12 | ], 13 | "rates": [ 14 | 0.5, 15 | 1.0, 16 | 1.5 17 | ], 18 | "img_rate_thr": 0.6, 19 | "iof_thr": 0.7, 20 | "no_padding": false, 21 | "padding_value": [ 22 | 104, 23 | 116, 24 | 124 25 | ], 26 | "save_dir": "data/split_ms_dota/test/", 27 | "save_ext": ".png" 28 | } 29 | -------------------------------------------------------------------------------- /tools/data/dota/split/split_configs/ms_train.json: -------------------------------------------------------------------------------- 1 | { 2 | "nproc": 10, 3 | "img_dirs": [ 4 | "data/DOTA/train/images/" 5 | ], 6 | "ann_dirs": [ 7 | "data/DOTA/train/labelTxt/" 8 | ], 9 | "sizes": [ 10 | 1024 11 | ], 12 | "gaps": [ 13 | 500 14 | ], 15 | "rates": [ 16 | 0.5, 17 | 1.0, 18 | 1.5 19 | ], 20 | "img_rate_thr": 0.6, 21 | "iof_thr": 0.7, 22 | "no_padding": false, 23 | "padding_value": [ 24 | 104, 25 | 116, 26 | 124 27 | ], 28 | "save_dir": "data/split_ms_dota/train/", 29 | "save_ext": ".png" 30 | } 31 | -------------------------------------------------------------------------------- /tools/data/dota/split/split_configs/ms_trainval.json: -------------------------------------------------------------------------------- 1 | { 2 | "nproc": 10, 3 | "img_dirs": [ 4 | "data/DOTA/train/images/", 5 | "data/DOTA/val/images/" 6 | ], 7 | "ann_dirs": [ 8 | "data/DOTA/train/labelTxt/", 9 | "data/DOTA/val/labelTxt/" 10 | ], 11 | "sizes": [ 12 | 1024 13 | ], 14 | "gaps": [ 15 | 500 16 | ], 17 | "rates": [ 18 | 0.5, 19 | 1.0, 20 | 1.5 21 | ], 22 | "img_rate_thr": 0.6, 23 | "iof_thr": 0.7, 24 | "no_padding": false, 25 | "padding_value": [ 26 | 104, 27 | 116, 28 | 124 29 | ], 30 | "save_dir": "data/split_ms_dota/trainval/", 31 | "save_ext": ".png" 32 | } 33 | -------------------------------------------------------------------------------- /tools/data/dota/split/split_configs/ms_val.json: -------------------------------------------------------------------------------- 1 | { 2 | "nproc": 10, 3 | "img_dirs": [ 4 | "data/DOTA/val/images/" 5 | ], 6 | "ann_dirs": [ 7 | "data/DOTA/val/labelTxt/" 8 | ], 9 | "sizes": [ 10 | 1024 11 | ], 12 | "gaps": [ 13 | 500 14 | ], 15 | "rates": [ 16 | 0.5, 17 | 1.0, 18 | 1.5 19 | ], 20 | "img_rate_thr": 0.6, 21 | "iof_thr": 0.7, 22 | "no_padding": false, 23 | "padding_value": [ 24 | 104, 25 | 116, 26 | 124 27 | ], 28 | "save_dir": "data/split_ms_dota/val/", 29 | "save_ext": ".png" 30 | } 31 | -------------------------------------------------------------------------------- /tools/data/dota/split/split_configs/ss_test.json: -------------------------------------------------------------------------------- 1 | { 2 | "nproc": 10, 3 | "img_dirs": [ 4 | "data/DOTA/test/images/" 5 | ], 6 | "ann_dirs": null, 7 | "sizes": [ 8 | 1024 9 | ], 10 | "gaps": [ 11 | 200 12 | ], 13 | "rates": [ 14 | 1.0 15 | ], 16 | "img_rate_thr": 0.6, 17 | "iof_thr": 0.7, 18 | "no_padding": false, 19 | "padding_value": [ 20 | 104, 21 | 116, 22 | 124 23 | ], 24 | "save_dir": "data/split_ss_dota/test/", 25 | "save_ext": ".png" 26 | } 27 | -------------------------------------------------------------------------------- /tools/data/dota/split/split_configs/ss_train.json: -------------------------------------------------------------------------------- 1 | { 2 | "nproc": 10, 3 | "img_dirs": [ 4 | "data/DOTA/train/images/" 5 | ], 6 | "ann_dirs": [ 7 | "data/DOTA/train/labelTxt/" 8 | ], 9 | "sizes": [ 10 | 1024 11 | ], 12 | "gaps": [ 13 | 200 14 | ], 15 | "rates": [ 16 | 1.0 17 | ], 18 | "img_rate_thr": 0.6, 19 | "iof_thr": 0.7, 20 | "no_padding": false, 21 | "padding_value": [ 22 | 104, 23 | 116, 24 | 124 25 | ], 26 | "save_dir": "data/split_ss_dota/train/", 27 | "save_ext": ".png" 28 | } 29 | -------------------------------------------------------------------------------- /tools/data/dota/split/split_configs/ss_trainval.json: -------------------------------------------------------------------------------- 1 | { 2 | "nproc": 10, 3 | "img_dirs": [ 4 | "data/DOTA/train/images/", 5 | "data/DOTA/val/images/" 6 | ], 7 | "ann_dirs": [ 8 | "data/DOTA/train/labelTxt/", 9 | "data/DOTA/val/labelTxt/" 10 | ], 11 | "sizes": [ 12 | 1024 13 | ], 14 | "gaps": [ 15 | 200 16 | ], 17 | "rates": [ 18 | 1.0 19 | ], 20 | "img_rate_thr": 0.6, 21 | "iof_thr": 0.7, 22 | "no_padding": false, 23 | "padding_value": [ 24 | 104, 25 | 116, 26 | 124 27 | ], 28 | "save_dir": "data/split_ss_dota/trainval/", 29 | "save_ext": ".png" 30 | } 31 | -------------------------------------------------------------------------------- /tools/data/dota/split/split_configs/ss_val.json: -------------------------------------------------------------------------------- 1 | { 2 | "nproc": 10, 3 | "img_dirs": [ 4 | "data/DOTA/val/images/" 5 | ], 6 | "ann_dirs": [ 7 | "data/DOTA/val/labelTxt/" 8 | ], 9 | "sizes": [ 10 | 1024 11 | ], 12 | "gaps": [ 13 | 200 14 | ], 15 | "rates": [ 16 | 1.0 17 | ], 18 | "img_rate_thr": 0.6, 19 | "iof_thr": 0.7, 20 | "no_padding": false, 21 | "padding_value": [ 22 | 104, 23 | 116, 24 | 124 25 | ], 26 | "save_dir": "data/split_ss_dota/val/", 27 | "save_ext": ".png" 28 | } 29 | -------------------------------------------------------------------------------- /tools/data/hrsc/README.md: -------------------------------------------------------------------------------- 1 | # Preparing HRSC Dataset 2 | 3 | 4 | 5 | ```bibtex 6 | @conference{hrsc, 7 | author = {Zikun Liu. and Liu Yuan. and Lubin Weng. and Yiping Yang.}, 8 | title = {A High Resolution Optical Satellite Image Dataset for Ship Recognition and Some New Baselines}, 9 | booktitle = {Proceedings of the 6th International Conference on Pattern Recognition Applications and Methods - ICPRAM,}, 10 | year = {2017}, 11 | pages = {324-331}, 12 | publisher = {SciTePress}, 13 | organization = {INSTICC}, 14 | doi = {10.5220/0006120603240331}, 15 | isbn = {978-989-758-222-6}, 16 | issn = {2184-4313}, 17 | } 18 | ``` 19 | 20 | ## Download HRSC dataset 21 | 22 | The HRSC dataset can be downloaded from [here](https://aistudio.baidu.com/aistudio/datasetdetail/54106). 23 | 24 | The data structure is as follows: 25 | 26 | ```none 27 | mmrotate 28 | ├── mmrotate 29 | ├── tools 30 | ├── configs 31 | ├── data 32 | │ ├── hrsc 33 | │ │ ├── FullDataSet 34 | │ │ │ ├─ AllImages 35 | │ │ │ ├─ Annotations 36 | │ │ │ ├─ LandMask 37 | │ │ │ ├─ Segmentations 38 | │ │ ├── ImageSets 39 | ``` 40 | 41 | ## Change base config 42 | 43 | Please change `data_root` in `configs/_base_/datasets/hrsc.py` to `data/hrsc/`. 44 | -------------------------------------------------------------------------------- /tools/data/hrsid/README.md: -------------------------------------------------------------------------------- 1 | # Preparing HRSID Dataset 2 | 3 | 4 | 5 | ```bibtex 6 | @ARTICLE{HRSID_2020, 7 | author={Wei, Shunjun and Zeng, Xiangfeng and Qu, Qizhe and Wang, Mou and Su, Hao and Shi, Jun}, 8 | journal={IEEE Access}, 9 | title={HRSID: A High-Resolution SAR Images Dataset for Ship Detection and Instance Segmentation}, 10 | year={2020}, 11 | volume={8}, 12 | pages={120234-120254}, 13 | } 14 | ``` 15 | 16 | ## Download HRSID dataset 17 | 18 | The HRSID dataset can be downloaded from [Google drive](https://drive.google.com/file/d/1BZTU8Gyg20wqHXtBPFzRazn_lEdvhsbE/view). 19 | 20 | The data structure is as follows: 21 | 22 | ```none 23 | mmrotate 24 | ├── mmrotate 25 | ├── tools 26 | ├── configs 27 | ├── data 28 | │ ├── HRSID_JPG 29 | │ │ ├── JPEGImages 30 | │ │ ├── annotations 31 | ``` 32 | 33 | ## Change base config 34 | 35 | Please change `data_root` in `configs/_base_/datasets/hrisd.py` to `data/HRSID_JPG/`. 36 | -------------------------------------------------------------------------------- /tools/data/rsdd/README.md: -------------------------------------------------------------------------------- 1 | # Preparing RSDD Dataset 2 | 3 | 4 | 5 | ```bibtex 6 | @ARTICLE{RSDD2022, 7 | author = {C. Xu, H. Su, J. Li, Y. Liu, L. Yao, L. Gao, W. Yan and T. Wang}, 8 | title = {RSDD-SAR: Rotated Ship Detection Dataset in SAR Images}, 9 | journal = {Journal of Radars}, 10 | month = {Sep.}, 11 | year = {2022}, 12 | volume={11}, 13 | number={R22007}, 14 | pages={581}, 15 | } 16 | ``` 17 | 18 | ## Download RSDD dataset 19 | 20 | The RSDD dataset can be downloaded from [Google drive](https://drive.google.com/file/d/1PJxr7Tbr_ZAzuG8MNloDa4mLaRYCD3qc/view?usp=sharing). 21 | 22 | The data structure is as follows: 23 | 24 | ```none 25 | mmrotate 26 | ├── mmrotate 27 | ├── tools 28 | ├── configs 29 | ├── data 30 | │ ├── rsdd 31 | │ │ ├── Annotations 32 | │ │ ├── ImageSets 33 | │ │ ├── JPEGImages 34 | │ │ ├── JPEGValidation 35 | ``` 36 | 37 | ## Change base config 38 | 39 | Please change `data_root` in `configs/_base_/datasets/rsdd.py` to `data/rsdd/`. 40 | -------------------------------------------------------------------------------- /tools/data/srsdd/README.md: -------------------------------------------------------------------------------- 1 | # Preparing SRSDD Dataset 2 | 3 | 4 | 5 | ```bibtex 6 | @ARTICLE{SRSDD2021, 7 | author = {S. Lei, D. Lu and X. Qiu}, 8 | title = {SRSDD-v1.0: A high-resolution SAR rotation ship 9 | detection dataset}, 10 | journal = {Remote Senseing}, 11 | month = {Dec.}, 12 | year = {2021}, 13 | volume={13}, 14 | number={24}, 15 | pages={5104}, 16 | } 17 | ``` 18 | 19 | ## Download SRSDD dataset 20 | 21 | The SRSDD dataset can be downloaded from [Google drive](https://drive.google.com/file/d/1QtCjih1ChOmG-TOPUTlsL3WbMh0L-1zp/view?usp=sharing). 22 | 23 | The data structure is as follows: 24 | 25 | ```none 26 | mmrotate 27 | ├── mmrotate 28 | ├── tools 29 | ├── configs 30 | ├── data 31 | │ ├── srsdd 32 | │ │ ├── train 33 | │ │ ├── test 34 | ``` 35 | 36 | ## Change base config 37 | 38 | Please change `data_root` in `configs/_base_/datasets/srsdd.py` to `data/srsdd/`. 39 | -------------------------------------------------------------------------------- /tools/data/ssdd/README.md: -------------------------------------------------------------------------------- 1 | # Preparing SSDD Dataset 2 | 3 | 4 | 5 | ```bibtex 6 | @ARTICLE{SSDD2021, 7 | author = {T. Zhang, X. Zhang, J. Li and X. Xu}, 8 | title = {SAR ship detection dataset (SSDD): Official release and comprehensive data analysis}, 9 | journal = {Remote Senseing}, 10 | month = {Sep.}, 11 | year = {2021} 12 | volume={13}, 13 | number={18}, 14 | pages={3690}, 15 | } 16 | ``` 17 | 18 | ## Download SSDD dataset 19 | 20 | The SSDD dataset can be downloaded from [Google drive](https://drive.google.com/file/d/1LmoHBk4xUvm0Zdtm8X7256dHigyFW4Nh/view?usp=sharing). 21 | 22 | The data structure is as follows: 23 | 24 | ```none 25 | mmrotate 26 | ├── mmrotate 27 | ├── tools 28 | ├── configs 29 | ├── data 30 | │ ├── ssdd 31 | │ │ ├── train 32 | │ │ ├── test 33 | │ │ │ ├── all 34 | │ │ │ ├── inshore 35 | │ │ │ ├── offshore 36 | ``` 37 | 38 | ## Change base config 39 | 40 | Please change `data_root` in `configs/_base_/datasets/ssdd.py` to `data/ssdd/`. 41 | -------------------------------------------------------------------------------- /tools/dist_test.sh: -------------------------------------------------------------------------------- 1 | #!/usr/bin/env bash 2 | 3 | CONFIG=$1 4 | CHECKPOINT=$2 5 | GPUS=$3 6 | NNODES=${NNODES:-1} 7 | NODE_RANK=${NODE_RANK:-0} 8 | PORT=${PORT:-29500} 9 | MASTER_ADDR=${MASTER_ADDR:-"127.0.0.1"} 10 | 11 | PYTHONPATH="$(dirname $0)/..":$PYTHONPATH \ 12 | python -m torch.distributed.launch \ 13 | --nnodes=$NNODES \ 14 | --node_rank=$NODE_RANK \ 15 | --master_addr=$MASTER_ADDR \ 16 | --nproc_per_node=$GPUS \ 17 | --master_port=$PORT \ 18 | $(dirname "$0")/test.py \ 19 | $CONFIG \ 20 | $CHECKPOINT \ 21 | --launcher pytorch \ 22 | ${@:4} 23 | -------------------------------------------------------------------------------- /tools/dist_train.sh: -------------------------------------------------------------------------------- 1 | #!/usr/bin/env bash 2 | 3 | CONFIG=$1 4 | GPUS=$2 5 | NNODES=${NNODES:-1} 6 | NODE_RANK=${NODE_RANK:-0} 7 | PORT=${PORT:-29500} 8 | MASTER_ADDR=${MASTER_ADDR:-"127.0.0.1"} 9 | 10 | PYTHONPATH="$(dirname $0)/..":$PYTHONPATH \ 11 | python -m torch.distributed.launch \ 12 | --nnodes=$NNODES \ 13 | --node_rank=$NODE_RANK \ 14 | --master_addr=$MASTER_ADDR \ 15 | --nproc_per_node=$GPUS \ 16 | --master_port=$PORT \ 17 | $(dirname "$0")/train.py \ 18 | $CONFIG \ 19 | --launcher pytorch ${@:3} 20 | -------------------------------------------------------------------------------- /tools/misc/print_config.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | import argparse 3 | 4 | from mmengine import Config, DictAction 5 | 6 | 7 | def parse_args(): 8 | """Parse arguments.""" 9 | parser = argparse.ArgumentParser(description='Print the whole config') 10 | parser.add_argument('config', help='config file path') 11 | parser.add_argument( 12 | '--cfg-options', 13 | nargs='+', 14 | action=DictAction, 15 | help='override some settings in the used config, the key-value pair ' 16 | 'in xxx=yyy format will be merged into config file. If the value to ' 17 | 'be overwritten is a list, it should be like key="[a,b]" or key=a,b ' 18 | 'It also allows nested list/tuple values, e.g. key="[(a,b),(c,d)]" ' 19 | 'Note that the quotation marks are necessary and that no white space ' 20 | 'is allowed.') 21 | args = parser.parse_args() 22 | 23 | return args 24 | 25 | 26 | def main(): 27 | """Print config.""" 28 | args = parse_args() 29 | 30 | cfg = Config.fromfile(args.config) 31 | 32 | if args.cfg_options is not None: 33 | cfg.merge_from_dict(args.cfg_options) 34 | print(f'Config:\n{cfg.pretty_text}') 35 | 36 | 37 | if __name__ == '__main__': 38 | main() 39 | -------------------------------------------------------------------------------- /tools/model_converters/publish_model.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) OpenMMLab. All rights reserved. 2 | import argparse 3 | import subprocess 4 | 5 | import torch 6 | 7 | 8 | def parse_args(): 9 | """Parse parameters.""" 10 | parser = argparse.ArgumentParser( 11 | description='Process a checkpoint to be published') 12 | parser.add_argument('in_file', help='input checkpoint filename') 13 | parser.add_argument('out_file', help='output checkpoint filename') 14 | args = parser.parse_args() 15 | return args 16 | 17 | 18 | def process_checkpoint(in_file, out_file): 19 | """Only inference related parameters are retained. 20 | 21 | Args: 22 | in_file (str): Filename of input checkpoint. 23 | out_file (str): Filename of output checkpoint. 24 | """ 25 | checkpoint = torch.load(in_file, map_location='cpu') 26 | # remove optimizer for smaller file size 27 | if 'optimizer' in checkpoint: 28 | del checkpoint['optimizer'] 29 | # if it is necessary to remove some sensitive data in checkpoint['meta'], 30 | # add the code here. 31 | if torch.__version__ >= '1.6': 32 | torch.save(checkpoint, out_file, _use_new_zipfile_serialization=False) 33 | else: 34 | torch.save(checkpoint, out_file) 35 | sha = subprocess.check_output(['sha256sum', out_file]).decode() 36 | if out_file.endswith('.pth'): 37 | out_file_name = out_file[:-4] 38 | else: 39 | out_file_name = out_file 40 | final_file = out_file_name + f'-{sha[:8]}.pth' 41 | subprocess.Popen(['mv', out_file, final_file]) 42 | 43 | 44 | def main(): 45 | """Main function of publish model.""" 46 | args = parse_args() 47 | process_checkpoint(args.in_file, args.out_file) 48 | 49 | 50 | if __name__ == '__main__': 51 | main() 52 | -------------------------------------------------------------------------------- /tools/slurm_test.sh: -------------------------------------------------------------------------------- 1 | #!/usr/bin/env bash 2 | 3 | set -x 4 | 5 | PARTITION=$1 6 | JOB_NAME=$2 7 | CONFIG=$3 8 | CHECKPOINT=$4 9 | GPUS=${GPUS:-8} 10 | GPUS_PER_NODE=${GPUS_PER_NODE:-8} 11 | CPUS_PER_TASK=${CPUS_PER_TASK:-5} 12 | PY_ARGS=${@:5} 13 | SRUN_ARGS=${SRUN_ARGS:-""} 14 | 15 | PYTHONPATH="$(dirname $0)/..":$PYTHONPATH \ 16 | srun -p ${PARTITION} \ 17 | --job-name=${JOB_NAME} \ 18 | --gres=gpu:${GPUS_PER_NODE} \ 19 | --ntasks=${GPUS} \ 20 | --ntasks-per-node=${GPUS_PER_NODE} \ 21 | --cpus-per-task=${CPUS_PER_TASK} \ 22 | --kill-on-bad-exit=1 \ 23 | ${SRUN_ARGS} \ 24 | python -u tools/test.py ${CONFIG} ${CHECKPOINT} --launcher="slurm" ${PY_ARGS} 25 | -------------------------------------------------------------------------------- /tools/slurm_train.sh: -------------------------------------------------------------------------------- 1 | #!/usr/bin/env bash 2 | 3 | set -x 4 | 5 | PARTITION=$1 6 | JOB_NAME=$2 7 | CONFIG=$3 8 | WORK_DIR=$4 9 | GPUS=${GPUS:-8} 10 | GPUS_PER_NODE=${GPUS_PER_NODE:-8} 11 | CPUS_PER_TASK=${CPUS_PER_TASK:-5} 12 | SRUN_ARGS=${SRUN_ARGS:-""} 13 | PY_ARGS=${@:5} 14 | 15 | PYTHONPATH="$(dirname $0)/..":$PYTHONPATH \ 16 | srun -p ${PARTITION} \ 17 | --job-name=${JOB_NAME} \ 18 | --gres=gpu:${GPUS_PER_NODE} \ 19 | --ntasks=${GPUS} \ 20 | --ntasks-per-node=${GPUS_PER_NODE} \ 21 | --cpus-per-task=${CPUS_PER_TASK} \ 22 | --kill-on-bad-exit=1 \ 23 | ${SRUN_ARGS} \ 24 | python -u tools/train.py ${CONFIG} --work-dir=${WORK_DIR} --launcher="slurm" ${PY_ARGS} 25 | --------------------------------------------------------------------------------