├── .gitignore ├── DOTA_devkit ├── ImgSplit_multi_process.py ├── ResultMerge_multi_process.py ├── SplitOnlyImage_multi_process.py ├── __init__.py ├── convert_dota_to_mmdet.py ├── dota-v1.5_evaluation_task1.py ├── dota-v1.5_evaluation_task2.py ├── dota_evaluation_task1.py ├── dota_evaluation_task2.py ├── dota_utils.py ├── hrsc2016_evaluation.py ├── polyiou │ ├── __init__.py │ ├── csrc │ │ ├── polyiou.cpp │ │ ├── polyiou.h │ │ └── polyiou.i │ ├── polyiou.py │ └── setup.py └── prepare_dota1_ms.py ├── README.md ├── Windows ├── README.md └── env.yml ├── configs ├── albu_example │ └── mask_rcnn_r50_fpn_1x.py ├── atss │ └── atss_r50_fpn_1x.py ├── cascade_mask_rcnn_r101_fpn_1x.py ├── cascade_mask_rcnn_r50_caffe_c4_1x.py ├── cascade_mask_rcnn_r50_fpn_1x.py ├── cascade_mask_rcnn_x101_32x4d_fpn_1x.py ├── cascade_mask_rcnn_x101_64x4d_fpn_1x.py ├── cascade_rcnn_r101_fpn_1x.py ├── cascade_rcnn_r50_caffe_c4_1x.py ├── cascade_rcnn_r50_fpn_1x.py ├── cascade_rcnn_x101_32x4d_fpn_1x.py ├── cascade_rcnn_x101_64x4d_fpn_1x.py ├── cityscapes │ ├── README.md │ ├── faster_rcnn_r50_fpn_1x_cityscapes.py │ └── mask_rcnn_r50_fpn_1x_cityscapes.py ├── dcn │ ├── README.md │ ├── cascade_mask_rcnn_dconv_c3-c5_r50_fpn_1x.py │ ├── cascade_rcnn_dconv_c3-c5_r50_fpn_1x.py │ ├── faster_rcnn_dconv_c3-c5_r50_fpn_1x.py │ ├── faster_rcnn_dconv_c3-c5_x101_32x4d_fpn_1x.py │ ├── faster_rcnn_dpool_r50_fpn_1x.py │ ├── faster_rcnn_mdconv_c3-c5_group4_r50_fpn_1x.py │ ├── faster_rcnn_mdconv_c3-c5_r50_fpn_1x.py │ ├── faster_rcnn_mdpool_r50_fpn_1x.py │ └── mask_rcnn_dconv_c3-c5_r50_fpn_1x.py ├── dota │ ├── cascade_s2anet_1s_r50_fpn_1x_dota.py │ ├── cascade_s2anet_2s_r50_fpn_1x_dota.py │ ├── faster_rcnn_hbb_obb_r50_fpn_1x_dota.py │ ├── retinanet_obb_r50_fpn_1x_dota.py │ └── s2anet_r50_fpn_1x_dota.py ├── double_heads │ └── dh_faster_rcnn_r50_fpn_1x.py ├── empirical_attention │ ├── README.md │ ├── faster_rcnn_r50_fpn_attention_0010_1x.py │ ├── faster_rcnn_r50_fpn_attention_0010_dcn_1x.py │ ├── faster_rcnn_r50_fpn_attention_1111_1x.py │ └── faster_rcnn_r50_fpn_attention_1111_dcn_1x.py ├── fast_mask_rcnn_r101_fpn_1x.py ├── fast_mask_rcnn_r50_caffe_c4_1x.py ├── fast_mask_rcnn_r50_fpn_1x.py ├── fast_rcnn_r101_fpn_1x.py ├── fast_rcnn_r50_caffe_c4_1x.py ├── fast_rcnn_r50_fpn_1x.py ├── faster_rcnn_ohem_r50_fpn_1x.py ├── faster_rcnn_r101_fpn_1x.py ├── faster_rcnn_r50_caffe_c4_1x.py ├── faster_rcnn_r50_fpn_1x.py ├── faster_rcnn_x101_32x4d_fpn_1x.py ├── faster_rcnn_x101_64x4d_fpn_1x.py ├── fcos │ ├── README.md │ ├── fcos_mstrain_640_800_r101_caffe_fpn_gn_2x_4gpu.py │ ├── fcos_mstrain_640_800_x101_64x4d_fpn_gn_2x.py │ └── fcos_r50_caffe_fpn_gn_1x_4gpu.py ├── foveabox │ ├── README.md │ ├── fovea_align_gn_ms_r101_fpn_4gpu_2x.py │ ├── fovea_align_gn_ms_r50_fpn_4gpu_2x.py │ ├── fovea_align_gn_r101_fpn_4gpu_2x.py │ ├── fovea_align_gn_r50_fpn_4gpu_2x.py │ └── fovea_r50_fpn_4gpu_1x.py ├── fp16 │ ├── faster_rcnn_r50_fpn_fp16_1x.py │ ├── mask_rcnn_r50_fpn_fp16_1x.py │ └── retinanet_r50_fpn_fp16_1x.py ├── gcnet │ ├── README.md │ ├── mask_rcnn_r16_gcb_c3-c5_r50_fpn_1x.py │ ├── mask_rcnn_r16_gcb_c3-c5_r50_fpn_syncbn_1x.py │ ├── mask_rcnn_r4_gcb_c3-c5_r50_fpn_1x.py │ ├── mask_rcnn_r4_gcb_c3-c5_r50_fpn_syncbn_1x.py │ └── mask_rcnn_r50_fpn_sbn_1x.py ├── ghm │ ├── README.md │ └── retinanet_ghm_r50_fpn_1x.py ├── gn+ws │ ├── README.md │ ├── faster_rcnn_r50_fpn_gn_ws_1x.py │ ├── mask_rcnn_r50_fpn_gn_ws_20_23_24e.py │ ├── mask_rcnn_r50_fpn_gn_ws_2x.py │ └── mask_rcnn_x101_32x4d_fpn_gn_ws_2x.py ├── gn │ ├── README.md │ ├── mask_rcnn_r101_fpn_gn_2x.py │ ├── mask_rcnn_r50_fpn_gn_2x.py │ └── mask_rcnn_r50_fpn_gn_contrib_2x.py ├── grid_rcnn │ ├── README.md │ ├── grid_rcnn_gn_head_r50_fpn_2x.py │ └── grid_rcnn_gn_head_x101_32x4d_fpn_2x.py ├── guided_anchoring │ ├── README.md │ ├── ga_fast_r50_caffe_fpn_1x.py │ ├── ga_faster_r50_caffe_fpn_1x.py │ ├── ga_faster_x101_32x4d_fpn_1x.py │ ├── ga_retinanet_r50_caffe_fpn_1x.py │ ├── ga_retinanet_x101_32x4d_fpn_1x.py │ ├── ga_rpn_r101_caffe_rpn_1x.py │ ├── ga_rpn_r50_caffe_fpn_1x.py │ └── ga_rpn_x101_32x4d_fpn_1x.py ├── hrnet │ ├── README.md │ ├── cascade_mask_rcnn_hrnetv2p_w32_20e.py │ ├── cascade_rcnn_hrnetv2p_w32_20e.py │ ├── faster_rcnn_hrnetv2p_w18_1x.py │ ├── faster_rcnn_hrnetv2p_w32_1x.py │ ├── faster_rcnn_hrnetv2p_w40_1x.py │ ├── fcos_hrnetv2p_w32_gn_1x_4gpu.py │ ├── htc_hrnetv2p_w32_20e.py │ ├── mask_rcnn_hrnetv2p_w18_1x.py │ └── mask_rcnn_hrnetv2p_w32_1x.py ├── hrsc2016 │ ├── cascade_s2anet_1s_r50_fpn_4x_hrsc2016.py │ ├── cascade_s2anet_2s_r50_fpn_3x_hrsc2016.py │ ├── retinanet_obb_r50_fpn_6x_hrsc2016.py │ ├── s2anet_r101_fpn_3x_hrsc2016.py │ └── s2anet_r50_fpn_3x_hrsc2016.py ├── htc │ ├── README.md │ ├── htc_dconv_c3-c5_mstrain_400_1400_x101_64x4d_fpn_20e.py │ ├── htc_r101_fpn_20e.py │ ├── htc_r50_fpn_1x.py │ ├── htc_r50_fpn_20e.py │ ├── htc_without_semantic_r50_fpn_1x.py │ ├── htc_x101_32x4d_fpn_20e_16gpu.py │ └── htc_x101_64x4d_fpn_20e_16gpu.py ├── libra_rcnn │ ├── README.md │ ├── libra_fast_rcnn_r50_fpn_1x.py │ ├── libra_faster_rcnn_r101_fpn_1x.py │ ├── libra_faster_rcnn_r50_fpn_1x.py │ ├── libra_faster_rcnn_x101_64x4d_fpn_1x.py │ └── libra_retinanet_r50_fpn_1x.py ├── mask_rcnn_r101_fpn_1x.py ├── mask_rcnn_r50_caffe_c4_1x.py ├── mask_rcnn_r50_fpn_1x.py ├── mask_rcnn_x101_32x4d_fpn_1x.py ├── mask_rcnn_x101_64x4d_fpn_1x.py ├── ms_rcnn │ ├── README.md │ ├── ms_rcnn_r101_caffe_fpn_1x.py │ ├── ms_rcnn_r50_caffe_fpn_1x.py │ └── ms_rcnn_x101_64x4d_fpn_1x.py ├── pascal_voc │ ├── README.md │ ├── faster_rcnn_r50_fpn_1x_voc0712.py │ ├── ssd300_voc.py │ └── ssd512_voc.py ├── reppoints │ ├── README.md │ ├── bbox_r50_grid_center_fpn_1x.py │ ├── bbox_r50_grid_fpn_1x.py │ ├── reppoints.png │ ├── reppoints_minmax_r50_fpn_1x.py │ ├── reppoints_moment_r101_dcn_fpn_2x.py │ ├── reppoints_moment_r101_dcn_fpn_2x_mt.py │ ├── reppoints_moment_r101_fpn_2x.py │ ├── reppoints_moment_r101_fpn_2x_mt.py │ ├── reppoints_moment_r50_fpn_1x.py │ ├── reppoints_moment_r50_fpn_2x.py │ ├── reppoints_moment_r50_fpn_2x_mt.py │ ├── reppoints_moment_x101_dcn_fpn_2x.py │ ├── reppoints_moment_x101_dcn_fpn_2x_mt.py │ └── reppoints_partial_minmax_r50_fpn_1x.py ├── retinanet_r101_fpn_1x.py ├── retinanet_r50_fpn_1x.py ├── retinanet_x101_32x4d_fpn_1x.py ├── retinanet_x101_64x4d_fpn_1x.py ├── rotated_iou │ ├── README.md │ ├── cascade_s2anet_2s_r50_fpn_1x_dota_iouloss.py │ └── retinanet_obb_r50_fpn_6x_hrsc2016_iouloss.py ├── rpn_r101_fpn_1x.py ├── rpn_r50_caffe_c4_1x.py ├── rpn_r50_fpn_1x.py ├── rpn_x101_32x4d_fpn_1x.py ├── rpn_x101_64x4d_fpn_1x.py ├── scratch │ ├── README.md │ ├── scratch_faster_rcnn_r50_fpn_gn_6x.py │ └── scratch_mask_rcnn_r50_fpn_gn_6x.py ├── ssd300_coco.py ├── ssd512_coco.py └── wider_face │ ├── README.md │ └── ssd300_wider_face.py ├── demo ├── cascade_s2anet.png ├── corruptions_sev_3.png ├── data_pipeline.png ├── demo.jpg ├── demo_inference.py ├── inference_demo.ipynb ├── network.png └── webcam_demo.py ├── docker └── Dockerfile ├── docs ├── CASCADE_S2ANET.md ├── DATA_PIPELINE.md ├── GETTING_STARTED.md ├── INSTALL.md ├── MODEL_ZOO.md ├── ROBUSTNESS_BENCHMARKING.md └── TECHNICAL_DETAILS.md ├── mmdet ├── __init__.py ├── apis │ ├── __init__.py │ ├── env.py │ ├── inference.py │ └── train.py ├── core │ ├── __init__.py │ ├── anchor │ │ ├── __init__.py │ │ ├── anchor_generator.py │ │ ├── anchor_generator_rotated.py │ │ ├── anchor_target.py │ │ ├── guided_anchor_target.py │ │ ├── point_generator.py │ │ └── point_target.py │ ├── bbox │ │ ├── __init__.py │ │ ├── assign_sampling.py │ │ ├── assigners │ │ │ ├── __init__.py │ │ │ ├── approx_max_iou_assigner.py │ │ │ ├── assign_result.py │ │ │ ├── base_assigner.py │ │ │ ├── max_iou_assigner.py │ │ │ └── point_assigner.py │ │ ├── bbox_target.py │ │ ├── bbox_target_rotated.py │ │ ├── builder.py │ │ ├── coder │ │ │ ├── __init__.py │ │ │ ├── base_bbox_coder.py │ │ │ ├── delta_xywh_bbox_coder.py │ │ │ ├── delta_xywha_bbox_coder.py │ │ │ └── pseudo_bbox_coder.py │ │ ├── iou_calculators │ │ │ ├── __init__.py │ │ │ ├── builder.py │ │ │ ├── iou2d_calculator.py │ │ │ └── iou2d_calculator_rotated.py │ │ ├── samplers │ │ │ ├── __init__.py │ │ │ ├── base_sampler.py │ │ │ ├── combined_sampler.py │ │ │ ├── instance_balanced_pos_sampler.py │ │ │ ├── iou_balanced_neg_sampler.py │ │ │ ├── ohem_sampler.py │ │ │ ├── pseudo_sampler.py │ │ │ ├── random_sampler.py │ │ │ ├── random_sampler_rotated.py │ │ │ └── sampling_result.py │ │ ├── transforms.py │ │ └── transforms_rotated.py │ ├── evaluation │ │ ├── __init__.py │ │ ├── bbox_overlaps.py │ │ ├── class_names.py │ │ ├── coco_utils.py │ │ ├── dota_utils.py │ │ ├── eval_hooks.py │ │ ├── mean_ap.py │ │ └── recall.py │ ├── fp16 │ │ ├── __init__.py │ │ ├── decorators.py │ │ ├── hooks.py │ │ └── utils.py │ ├── mask │ │ ├── __init__.py │ │ ├── mask_target.py │ │ └── utils.py │ ├── post_processing │ │ ├── __init__.py │ │ ├── bbox_nms.py │ │ ├── bbox_nms_rotated.py │ │ ├── merge_augs.py │ │ └── merge_augs_rotated.py │ └── utils │ │ ├── __init__.py │ │ ├── dist_utils.py │ │ └── misc.py ├── datasets │ ├── __init__.py │ ├── builder.py │ ├── cityscapes.py │ ├── coco.py │ ├── custom.py │ ├── dataset_wrappers.py │ ├── dota.py │ ├── hrsc2016.py │ ├── loader │ │ ├── __init__.py │ │ ├── build_loader.py │ │ └── sampler.py │ ├── pipelines │ │ ├── __init__.py │ │ ├── compose.py │ │ ├── formating.py │ │ ├── loading.py │ │ ├── test_aug.py │ │ ├── transforms.py │ │ └── transforms_rotated.py │ ├── registry.py │ ├── voc.py │ ├── wider_face.py │ └── xml_style.py ├── models │ ├── __init__.py │ ├── anchor_heads │ │ ├── __init__.py │ │ ├── anchor_head.py │ │ ├── fcos_head.py │ │ ├── fovea_head.py │ │ ├── fsaf_head.py │ │ ├── ga_retina_head.py │ │ ├── ga_rpn_head.py │ │ ├── guided_anchor_head.py │ │ ├── reppoints_head.py │ │ ├── retina_head.py │ │ ├── rpn_head.py │ │ └── ssd_head.py │ ├── anchor_heads_rotated │ │ ├── __init__.py │ │ ├── anchor_head_rotated.py │ │ ├── cascade_s2anet_head.py │ │ ├── retina_head_rotated.py │ │ └── s2anet_head.py │ ├── backbones │ │ ├── __init__.py │ │ ├── hrnet.py │ │ ├── resnet.py │ │ ├── resnext.py │ │ └── ssd_vgg.py │ ├── bbox_heads │ │ ├── __init__.py │ │ ├── bbox_head.py │ │ ├── convfc_bbox_head.py │ │ └── double_bbox_head.py │ ├── bbox_heads_rotated │ │ ├── __init__.py │ │ ├── bbox_head_rotated.py │ │ ├── convfc_bbox_head_rotated.py │ │ └── double_bbox_head_rotated.py │ ├── builder.py │ ├── detectors │ │ ├── __init__.py │ │ ├── base.py │ │ ├── cascade_rcnn.py │ │ ├── cascade_s2anet.py │ │ ├── double_head_rcnn.py │ │ ├── fast_rcnn.py │ │ ├── faster_rcnn.py │ │ ├── faster_rcnn_hbb_obb.py │ │ ├── fcos.py │ │ ├── fovea.py │ │ ├── grid_rcnn.py │ │ ├── htc.py │ │ ├── mask_rcnn.py │ │ ├── mask_scoring_rcnn.py │ │ ├── reppoints_detector.py │ │ ├── retinanet.py │ │ ├── rpn.py │ │ ├── s2anet.py │ │ ├── single_stage.py │ │ ├── test_mixins.py │ │ └── two_stage.py │ ├── losses │ │ ├── __init__.py │ │ ├── accuracy.py │ │ ├── balanced_l1_loss.py │ │ ├── cross_entropy_loss.py │ │ ├── focal_loss.py │ │ ├── ghm_loss.py │ │ ├── iou_loss.py │ │ ├── mse_loss.py │ │ ├── rotated_iou_loss.py │ │ ├── smooth_l1_loss.py │ │ └── utils.py │ ├── mask_heads │ │ ├── __init__.py │ │ ├── fcn_mask_head.py │ │ ├── fused_semantic_head.py │ │ ├── grid_head.py │ │ ├── htc_mask_head.py │ │ └── maskiou_head.py │ ├── necks │ │ ├── __init__.py │ │ ├── bfp.py │ │ ├── fpn.py │ │ └── hrfpn.py │ ├── plugins │ │ ├── __init__.py │ │ ├── generalized_attention.py │ │ └── non_local.py │ ├── registry.py │ ├── roi_extractors │ │ ├── __init__.py │ │ ├── single_level.py │ │ └── single_level_rotated.py │ ├── shared_heads │ │ ├── __init__.py │ │ └── res_layer.py │ └── utils │ │ ├── __init__.py │ │ ├── conv_module.py │ │ ├── conv_ws.py │ │ ├── norm.py │ │ ├── scale.py │ │ └── weight_init.py ├── ops │ ├── __init__.py │ ├── box_iou_rotated │ │ ├── __init__.py │ │ └── src │ │ │ ├── box_iou_rotated.h │ │ │ ├── box_iou_rotated_cpu.cpp │ │ │ ├── box_iou_rotated_cuda.cu │ │ │ └── box_iou_rotated_utils.h │ ├── box_iou_rotated_diff │ │ ├── __init__.py │ │ ├── box_intersection_2d.py │ │ ├── box_iou_rotated_diff.py │ │ └── src │ │ │ ├── cuda_utils.h │ │ │ ├── sort_vert.cpp │ │ │ ├── sort_vert.h │ │ │ ├── sort_vert_kernel.cu │ │ │ └── utils.h │ ├── context_block.py │ ├── dcn │ │ ├── __init__.py │ │ ├── deform_conv.py │ │ ├── deform_pool.py │ │ └── src │ │ │ ├── deform_conv_cuda.cpp │ │ │ ├── deform_conv_cuda_kernel.cu │ │ │ ├── deform_pool_cuda.cpp │ │ │ └── deform_pool_cuda_kernel.cu │ ├── masked_conv │ │ ├── __init__.py │ │ ├── masked_conv.py │ │ └── src │ │ │ ├── masked_conv2d_cuda.cpp │ │ │ └── masked_conv2d_kernel.cu │ ├── ml_nms_rotated │ │ ├── __init__.py │ │ └── src │ │ │ ├── box_iou_rotated_utils.h │ │ │ ├── nms_rotated.h │ │ │ ├── nms_rotated_cpu.cpp │ │ │ └── nms_rotated_cuda.cu │ ├── nms │ │ ├── __init__.py │ │ ├── nms_wrapper.py │ │ └── src │ │ │ ├── nms_cpu.cpp │ │ │ ├── nms_cuda.cpp │ │ │ ├── nms_kernel.cu │ │ │ └── soft_nms_cpu.pyx │ ├── nms_rotated │ │ ├── __init__.py │ │ └── src │ │ │ ├── box_iou_rotated_utils.h │ │ │ ├── nms_rotated.h │ │ │ ├── nms_rotated_cpu.cpp │ │ │ └── nms_rotated_cuda.cu │ ├── orn │ │ ├── __init__.py │ │ ├── functions │ │ │ ├── __init__.py │ │ │ ├── active_rotating_filter.py │ │ │ ├── rotation_invariant_encoding.py │ │ │ └── rotation_invariant_pooling.py │ │ ├── modules │ │ │ ├── ORConv.py │ │ │ └── __init__.py │ │ └── src │ │ │ ├── ActiveRotatingFilter.h │ │ │ ├── RotationInvariantEncoding.h │ │ │ ├── cpu │ │ │ ├── ActiveRotatingFilter_cpu.cpp │ │ │ ├── RotationInvariantEncoding_cpu.cpp │ │ │ └── vision.h │ │ │ ├── cuda │ │ │ ├── ActiveRotatingFilter_cuda.cu │ │ │ ├── RotationInvariantEncoding_cuda.cu │ │ │ └── vision.h │ │ │ └── vision.cpp │ ├── roi_align │ │ ├── __init__.py │ │ ├── gradcheck.py │ │ ├── roi_align.py │ │ └── src │ │ │ ├── roi_align_cuda.cpp │ │ │ └── roi_align_kernel.cu │ ├── roi_align_rotated │ │ ├── __init__.py │ │ ├── roi_align_rotated.py │ │ └── src │ │ │ ├── ROIAlignRotated.h │ │ │ ├── ROIAlignRotated_cpu.cpp │ │ │ └── ROIAlignRotated_cuda.cu │ ├── roi_pool │ │ ├── __init__.py │ │ ├── gradcheck.py │ │ ├── roi_pool.py │ │ └── src │ │ │ ├── roi_pool_cuda.cpp │ │ │ └── roi_pool_kernel.cu │ └── sigmoid_focal_loss │ │ ├── __init__.py │ │ ├── sigmoid_focal_loss.py │ │ └── src │ │ ├── sigmoid_focal_loss.cpp │ │ └── sigmoid_focal_loss_cuda.cu └── utils │ ├── __init__.py │ ├── flops_counter.py │ └── registry.py ├── requirements.txt ├── setup.py └── tools ├── analyze_logs.py ├── coco_error_analysis.py ├── coco_eval.py ├── convert_datasets └── pascal_voc.py ├── convert_model.py ├── detectron2pytorch.py ├── dist_test.sh ├── dist_train.sh ├── get_flops.py ├── publish_model.py ├── robustness_eval.py ├── slurm_test.sh ├── slurm_train.sh ├── slurm_train.slurm ├── test.py ├── test_robustness.py ├── train.py ├── upgrade_model_version.py └── voc_eval.py /.gitignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | **/__pycache__/ 4 | *.py[cod] 5 | *$py.class 6 | 7 | # C extensions 8 | *.so 9 | 10 | # Distribution / packaging 11 | .Python 12 | build/ 13 | develop-eggs/ 14 | dist/ 15 | downloads/ 16 | eggs/ 17 | .eggs/ 18 | lib/ 19 | lib64/ 20 | parts/ 21 | sdist/ 22 | var/ 23 | wheels/ 24 | *.egg-info/ 25 | .installed.cfg 26 | *.egg 27 | MANIFEST 28 | 29 | # PyInstaller 30 | # Usually these files are written by a python script from a template 31 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 32 | *.manifest 33 | *.spec 34 | 35 | # Installer logs 36 | pip-log.txt 37 | pip-delete-this-directory.txt 38 | 39 | # Unit test / coverage reports 40 | htmlcov/ 41 | .tox/ 42 | .coverage 43 | .coverage.* 44 | .cache 45 | nosetests.xml 46 | coverage.xml 47 | *.cover 48 | .hypothesis/ 49 | .pytest_cache/ 50 | 51 | # Translations 52 | *.mo 53 | *.pot 54 | 55 | # Django stuff: 56 | *.log 57 | local_settings.py 58 | db.sqlite3 59 | 60 | # Flask stuff: 61 | instance/ 62 | .webassets-cache 63 | 64 | # Scrapy stuff: 65 | .scrapy 66 | 67 | # Sphinx documentation 68 | docs/_build/ 69 | 70 | # PyBuilder 71 | target/ 72 | 73 | # Jupyter Notebook 74 | .ipynb_checkpoints 75 | 76 | # pyenv 77 | .python-version 78 | 79 | # celery beat schedule file 80 | celerybeat-schedule 81 | 82 | # SageMath parsed files 83 | *.sage.py 84 | 85 | # Environments 86 | .env 87 | .venv 88 | env/ 89 | venv/ 90 | ENV/ 91 | env.bak/ 92 | venv.bak/ 93 | 94 | # Spyder project settings 95 | .spyderproject 96 | .spyproject 97 | 98 | # Rope project settings 99 | .ropeproject 100 | 101 | # mkdocs documentation 102 | /site 103 | 104 | # mypy 105 | .mypy_cache/ 106 | 107 | # cython generated cpp 108 | mmdet/ops/nms/src/soft_nms_cpu.cpp 109 | mmdet/version.py 110 | data 111 | .vscode 112 | .idea 113 | 114 | # custom 115 | *.pkl 116 | *.pkl.json 117 | *.log.json 118 | work_dirs/ 119 | 120 | # Pytorch 121 | *.pth 122 | demo_results 123 | exps 124 | DOTA_devkit/polyiou/csrc/polyiou_wrap.cxx 125 | DOTA_devkit/polyiou/csrc/polyiou.py 126 | *_bak -------------------------------------------------------------------------------- /DOTA_devkit/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/csuhan/s2anet/e463c2bb17e161a6909e3d83410b49697dce770a/DOTA_devkit/__init__.py -------------------------------------------------------------------------------- /DOTA_devkit/polyiou/__init__.py: -------------------------------------------------------------------------------- 1 | from . import polyiou 2 | 3 | __all__ = ['polyiou'] -------------------------------------------------------------------------------- /DOTA_devkit/polyiou/csrc/polyiou.h: -------------------------------------------------------------------------------- 1 | // 2 | // Created by dingjian on 18-2-3. 3 | // 4 | 5 | #ifndef POLYIOU_POLYIOU_H 6 | #define POLYIOU_POLYIOU_H 7 | 8 | #include 9 | double iou_poly(std::vector p, std::vector q); 10 | #endif //POLYIOU_POLYIOU_H 11 | -------------------------------------------------------------------------------- /DOTA_devkit/polyiou/csrc/polyiou.i: -------------------------------------------------------------------------------- 1 | %module polyiou 2 | %include "std_vector.i" 3 | 4 | namespace std { 5 | %template(VectorDouble) vector; 6 | }; 7 | 8 | %{ 9 | #define SWIG_FILE_WITH_INIT 10 | #include 11 | #include 12 | #include 13 | #include 14 | 15 | #include "polyiou.h" 16 | %} 17 | 18 | %include "polyiou.h" 19 | 20 | -------------------------------------------------------------------------------- /DOTA_devkit/polyiou/setup.py: -------------------------------------------------------------------------------- 1 | from distutils.core import setup, Extension 2 | 3 | polyiou_module = Extension( 4 | '_polyiou', 5 | sources=['csrc/polyiou_wrap.cxx', 'csrc/polyiou.cpp']) 6 | setup(name='polyiou', 7 | version='0.1', 8 | author="SWIG Docs", 9 | description="""Simple swig example from docs""", 10 | ext_modules=[polyiou_module], 11 | py_modules=["polyiou"], 12 | ) 13 | -------------------------------------------------------------------------------- /DOTA_devkit/prepare_dota1_ms.py: -------------------------------------------------------------------------------- 1 | import os 2 | import os.path as osp 3 | 4 | from DOTA_devkit.ImgSplit_multi_process import splitbase as splitbase_trainval 5 | from DOTA_devkit.SplitOnlyImage_multi_process import splitbase as splitbase_test 6 | from DOTA_devkit.convert_dota_to_mmdet import convert_dota_to_mmdet 7 | 8 | 9 | def mkdir_if_not_exists(path): 10 | if not osp.exists(path): 11 | os.mkdir(path) 12 | 13 | 14 | def prepare_multi_scale_data(src_path, dst_path, gap=200, subsize=1024, scales=[0.5, 1.0, 1.5], num_process=32): 15 | dst_trainval_path = osp.join(dst_path, 'trainval_split') 16 | dst_test_base_path = osp.join(dst_path, 'test_split') 17 | dst_test_path = osp.join(dst_path, 'test_split/images') 18 | # make dst path if not exist 19 | mkdir_if_not_exists(dst_path) 20 | mkdir_if_not_exists(dst_trainval_path) 21 | mkdir_if_not_exists(dst_test_base_path) 22 | mkdir_if_not_exists(dst_test_path) 23 | 24 | # split train data 25 | print('split train data') 26 | split_train = splitbase_trainval(osp.join(src_path, 'train'), dst_trainval_path, 27 | gap=gap, subsize=subsize, num_process=num_process) 28 | for scale in scales: 29 | split_train.splitdata(scale) 30 | print('split val data') 31 | # split val data 32 | split_val = splitbase_trainval(osp.join(src_path, 'val'), dst_trainval_path, 33 | gap=gap, subsize=subsize, num_process=num_process) 34 | for scale in scales: 35 | split_val.splitdata(scale) 36 | # split test data 37 | print('split test data') 38 | split_test = splitbase_test(osp.join(src_path, 'test/images'), dst_test_path, 39 | gap=gap, subsize=subsize, num_process=num_process) 40 | for scale in scales: 41 | split_test.splitdata(scale) 42 | 43 | convert_dota_to_mmdet(dst_trainval_path, 44 | osp.join(dst_trainval_path, 'trainval1024.pkl')) 45 | convert_dota_to_mmdet(dst_test_base_path, 46 | osp.join(dst_test_base_path, 'test1024.pkl'), trainval=False) 47 | print('done!') 48 | 49 | 50 | if __name__ == '__main__': 51 | prepare_multi_scale_data('/data/hjm/dota', '/data/hjm/dota_1024', gap=200, subsize=1024, scales=[1.0], 52 | num_process=32) 53 | -------------------------------------------------------------------------------- /Windows/env.yml: -------------------------------------------------------------------------------- 1 | name: s2anet 2 | channels: 3 | - pytorch 4 | - defaults 5 | dependencies: 6 | - blas=1.0=mkl 7 | - ca-certificates=2023.08.22=haa95532_0 8 | - certifi=2022.12.7=py37haa95532_0 9 | - cudatoolkit=11.3.1=h59b6b97_2 10 | - flit-core=3.6.0=pyhd3eb1b0_0 11 | - freetype=2.12.1=ha860e81_0 12 | - giflib=5.2.1=h8cc25b3_3 13 | - intel-openmp=2021.4.0=haa95532_3556 14 | - jpeg=9e=h2bbff1b_1 15 | - lerc=3.0=hd77b12b_0 16 | - libdeflate=1.17=h2bbff1b_1 17 | - libpng=1.6.39=h8cc25b3_0 18 | - libtiff=4.5.1=hd77b12b_0 19 | - libuv=1.44.2=h2bbff1b_0 20 | - libwebp=1.3.2=hbc33d0d_0 21 | - libwebp-base=1.3.2=h2bbff1b_0 22 | - lz4-c=1.9.4=h2bbff1b_0 23 | - mkl=2021.4.0=haa95532_640 24 | - mkl-service=2.4.0=py37h2bbff1b_0 25 | - mkl_fft=1.3.1=py37h277e83a_0 26 | - mkl_random=1.2.2=py37hf11a4ad_0 27 | - numpy=1.21.5=py37h7a0a035_3 28 | - numpy-base=1.21.5=py37hca35cd5_3 29 | - openssl=1.1.1w=h2bbff1b_0 30 | - pillow=9.3.0=py37hdc2b20a_1 31 | - pip=22.3.1=py37haa95532_0 32 | - python=3.7.16=h6244533_0 33 | - pytorch=1.10.1=py3.7_cuda11.3_cudnn8_0 34 | - pytorch-mutex=1.0=cuda 35 | - setuptools=65.6.3=py37haa95532_0 36 | - six=1.16.0=pyhd3eb1b0_1 37 | - sqlite=3.41.2=h2bbff1b_0 38 | - tk=8.6.12=h2bbff1b_0 39 | - torchvision=0.11.2=py37_cu113 40 | - typing_extensions=4.4.0=py37haa95532_0 41 | - vc=14.2=h21ff451_1 42 | - vs2015_runtime=14.27.29016=h5e58377_2 43 | - wheel=0.38.4=py37haa95532_0 44 | - wincertstore=0.2=py37haa95532_2 45 | - xz=5.4.2=h8cc25b3_0 46 | - zlib=1.2.13=h8cc25b3_0 47 | - zstd=1.5.5=hd43e919_0 48 | - pip: 49 | - addict==2.4.0 50 | - albumentations==1.3.1 51 | - charset-normalizer==3.3.2 52 | - cycler==0.11.0 53 | - cython==3.0.5 54 | - fonttools==4.38.0 55 | - idna==3.4 56 | - imagecorruptions==1.1.2 57 | - imageio==2.31.2 58 | - joblib==1.3.2 59 | - kiwisolver==1.4.5 60 | - matplotlib==3.5.3 61 | - mmcv==0.2.14 62 | - networkx==2.6.3 63 | - opencv-python==4.8.1.78 64 | - opencv-python-headless==4.8.1.78 65 | - packaging==23.2 66 | - pandas==1.3.5 67 | - pycocotools==2.0.7 68 | - pyparsing==3.1.1 69 | - python-dateutil==2.8.2 70 | - pytz==2023.3.post1 71 | - pywavelets==1.3.0 72 | - pyyaml==6.0.1 73 | - qudida==0.0.4 74 | - requests==2.31.0 75 | - scikit-image==0.19.3 76 | - scikit-learn==1.0.2 77 | - scipy==1.7.3 78 | - shapely==2.0.2 79 | - terminaltables==3.1.10 80 | - threadpoolctl==3.1.0 81 | - tifffile==2021.11.2 82 | - urllib3==2.0.7 83 | -------------------------------------------------------------------------------- /configs/cityscapes/README.md: -------------------------------------------------------------------------------- 1 | ## Common settings 2 | 3 | - All baselines were trained using 8 GPU with a batch size of 8 (1 images per GPU) using the [linear scaling rule](https://arxiv.org/abs/1706.02677) to scale the learning rate. 4 | - All models were trained on `cityscapes_train`, and tested on `cityscapes_val`. 5 | - 1x training schedule indicates 64 epochs which corresponds to slightly less than the 24k iterations reported in the original schedule from the [Mask R-CNN paper](https://arxiv.org/abs/1703.06870) 6 | - All pytorch-style pretrained backbones on ImageNet are from PyTorch model zoo. 7 | 8 | 9 | ## Baselines 10 | 11 | Download links and more models with different backbones and training schemes will be added to the model zoo. 12 | 13 | 14 | ### Faster R-CNN 15 | 16 | | Backbone | Style | Lr schd | Scale | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | Download | 17 | | :-------------: | :-----: | :-----: | :---: | :------: | :-----------------: | :------------: | :----: | :------: | 18 | | R-50-FPN | pytorch | 1x | 800-1024 | 4.9 | 0.345 | 8.8 | 36.0 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/cityscapes/faster_rcnn_r50_fpn_1x_city_20190727-7b9c0534.pth) | 19 | 20 | ### Mask R-CNN 21 | 22 | | Backbone | Style | Lr schd | Scale | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | mask AP | Download | 23 | | :-------------: | :-----: | :-----: | :------: | :------: | :-----------------: | :------------: | :----: | :-----: | :------: | 24 | | R-50-FPN | pytorch | 1x | 800-1024 | 4.9 | 0.609 | 2.5 | 37.4 | 32.5 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/cityscapes/mask_rcnn_r50_fpn_1x_city_20190727-9b3c56a5.pth) | 25 | 26 | **Notes:** 27 | - In the original paper, the mask AP of Mask R-CNN R-50-FPN is 31.5. 28 | 29 | -------------------------------------------------------------------------------- /configs/empirical_attention/README.md: -------------------------------------------------------------------------------- 1 | # An Empirical Study of Spatial Attention Mechanisms in Deep Networks 2 | 3 | ## Introduction 4 | 5 | ``` 6 | @article{zhu2019empirical, 7 | title={An Empirical Study of Spatial Attention Mechanisms in Deep Networks}, 8 | author={Zhu, Xizhou and Cheng, Dazhi and Zhang, Zheng and Lin, Stephen and Dai, Jifeng}, 9 | journal={arXiv preprint arXiv:1904.05873}, 10 | year={2019} 11 | } 12 | ``` 13 | 14 | 15 | ## Results and Models 16 | 17 | | Backbone | Attention Component | DCN | Lr schd | box AP | Download | 18 | |:---------:|:-------------------:|:----:|:-------:|:------:|:--------:| 19 | | R-50 | 1111 | N | 1x | 38.6 | - | 20 | | R-50 | 0010 | N | 1x | 38.2 | - | 21 | | R-50 | 1111 | Y | 1x | 41.0 | - | 22 | | R-50 | 0010 | Y | 1x | 40.8 | - | 23 | 24 | -------------------------------------------------------------------------------- /configs/fcos/README.md: -------------------------------------------------------------------------------- 1 | # FCOS: Fully Convolutional One-Stage Object Detection 2 | 3 | ## Introduction 4 | 5 | ``` 6 | @article{tian2019fcos, 7 | title={FCOS: Fully Convolutional One-Stage Object Detection}, 8 | author={Tian, Zhi and Shen, Chunhua and Chen, Hao and He, Tong}, 9 | journal={arXiv preprint arXiv:1904.01355}, 10 | year={2019} 11 | } 12 | ``` 13 | 14 | ## Results and Models 15 | 16 | | Backbone | Style | GN | MS train | Lr schd | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | Download | 17 | |:---------:|:-------:|:-------:|:-------:|:-------:|:--------:|:-------------------:|:--------------:|:------:|:--------:| 18 | | R-50 | caffe | N | N | 1x | 5.5 | 0.373 | 13.7 | 35.7 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/fcos/fcos_r50_caffe_fpn_1x_4gpu_20190516-a7cac5ff.pth) | 19 | | R-50 | caffe | Y | N | 1x | 6.9 | 0.396 | 13.6 | 36.7 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/fcos/fcos_r50_caffe_fpn_gn_1x_4gpu_20190516-9f253a93.pth) | 20 | | R-50 | caffe | Y | N | 2x | - | - | - | 36.9 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/fcos/fcos_r50_caffe_fpn_gn_2x_4gpu_20190516_-93484354.pth) | 21 | | R-101 | caffe | Y | N | 1x | 10.4 | 0.558 | 11.6 | 39.1 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/fcos/fcos_r101_caffe_fpn_gn_1x_4gpu_20190516-e4889733.pth) | 22 | | R-101 | caffe | Y | N | 2x | - | - | - | 39.1 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/fcos/fcos_r101_caffe_fpn_gn_2x_4gpu_20190516-c03af97b.pth) | 23 | 24 | 25 | | Backbone | Style | GN | MS train | Lr schd | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | Download | 26 | |:---------:|:-------:|:-------:|:-------:|:-------:|:--------:|:-------------------:|:--------------:|:------:|:--------:| 27 | | R-50 | caffe | Y | Y | 2x | - | - | - | 38.7 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/fcos/fcos_mstrain_640_800_r50_caffe_fpn_gn_2x_4gpu_20190516-f7329d80.pth) | 28 | | R-101 | caffe | Y | Y | 2x | - | - | - | 40.8 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/fcos/fcos_mstrain_640_800_r101_caffe_fpn_gn_2x_4gpu_20190516-42e6f62d.pth) | 29 | | X-101 | caffe | Y | Y | 2x | 9.7 | 0.892 | 7.0 | 42.8 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/fcos/fcos_mstrain_640_800_x101_64x4d_fpn_gn_2x_20190516-a36c0872.pth) | 30 | 31 | **Notes:** 32 | - To be consistent with the author's implementation, we use 4 GPUs with 4 images/GPU for R-50 and R-101 models, and 8 GPUs with 2 image/GPU for X-101 models. 33 | - The X-101 backbone is X-101-64x4d. 34 | -------------------------------------------------------------------------------- /configs/ghm/README.md: -------------------------------------------------------------------------------- 1 | # Gradient Harmonized Single-stage Detector 2 | 3 | ## Introduction 4 | 5 | ``` 6 | @inproceedings{li2019gradient, 7 | title={Gradient Harmonized Single-stage Detector}, 8 | author={Li, Buyu and Liu, Yu and Wang, Xiaogang}, 9 | booktitle={AAAI Conference on Artificial Intelligence}, 10 | year={2019} 11 | } 12 | ``` 13 | 14 | ## Results and Models 15 | 16 | | Backbone | Style | Lr schd | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | Download | 17 | | :-------------: | :-----: | :-----: | :------: | :-----------------: | :------------: | :----: | :------: | 18 | | R-50-FPN | pytorch | 1x | 3.9 | 0.500 | 9.4 | 36.9 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/ghm/retinanet_ghm_r50_fpn_1x_20190608-b9aa5862.pth) | 19 | | R-101-FPN | pytorch | 1x | 5.8 | 0.625 | 8.5 | 39.0 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/ghm/retinanet_ghm_r101_fpn_1x_20190608-b885b74a.pth) | 20 | | X-101-32x4d-FPN | pytorch | 1x | 7.0 | 0.818 | 7.6 | 40.5 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/ghm/retinanet_ghm_x101_32x4d_fpn_1x_20190608-ed295d22.pth) | 21 | | X-101-64x4d-FPN | pytorch | 1x | 9.9 | 1.191 | 6.1 | 41.6 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/ghm/retinanet_ghm_x101_64x4d_fpn_1x_20190608-7f2037ce.pth) | -------------------------------------------------------------------------------- /configs/gn/README.md: -------------------------------------------------------------------------------- 1 | # Group Normalization 2 | 3 | ## Introduction 4 | 5 | ``` 6 | @inproceedings{wu2018group, 7 | title={Group Normalization}, 8 | author={Wu, Yuxin and He, Kaiming}, 9 | booktitle={Proceedings of the European Conference on Computer Vision (ECCV)}, 10 | year={2018} 11 | } 12 | ``` 13 | 14 | ## Results and Models 15 | 16 | | Backbone | model | Lr schd | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | mask AP | Download | 17 | |:-------------:|:----------:|:-------:|:--------:|:-------------------:|:--------------:|:------:|:-------:|:--------:| 18 | | R-50-FPN (d) | Mask R-CNN | 2x | 7.2 | 0.806 | 5.4 | 39.8 | 36.1 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/gn/mask_rcnn_r50_fpn_gn_2x_20180113-86832cf2.pth) | 19 | | R-50-FPN (d) | Mask R-CNN | 3x | 7.2 | 0.806 | 5.4 | 40.1 | 36.4 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/gn/mask_rcnn_r50_fpn_gn_3x_20180113-8e82f48d.pth) | 20 | | R-101-FPN (d) | Mask R-CNN | 2x | 9.9 | 0.970 | 4.8 | 41.5 | 37.0 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/gn/mask_rcnn_r101_fpn_gn_2x_20180113-9598649c.pth) | 21 | | R-101-FPN (d) | Mask R-CNN | 3x | 9.9 | 0.970 | 4.8 | 41.6 | 37.3 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/gn/mask_rcnn_r101_fpn_gn_3x_20180113-a14ffb96.pth) | 22 | | R-50-FPN (c) | Mask R-CNN | 2x | 7.2 | 0.806 | 5.4 | 39.7 | 35.9 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/gn/mask_rcnn_r50_fpn_gn_contrib_2x_20180113-ec93305c.pth) | 23 | | R-50-FPN (c) | Mask R-CNN | 3x | 7.2 | 0.806 | 5.4 | 40.0 | 36.2 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/gn/mask_rcnn_r50_fpn_gn_contrib_3x_20180113-9d230cab.pth) | 24 | 25 | **Notes:** 26 | - (d) means pretrained model converted from Detectron, and (c) means the contributed model pretrained by [@thangvubk](https://github.com/thangvubk). 27 | - The `3x` schedule is epoch [28, 34, 36]. 28 | - **Memory, Train/Inf time is outdated.** -------------------------------------------------------------------------------- /configs/grid_rcnn/README.md: -------------------------------------------------------------------------------- 1 | # Grid R-CNN 2 | 3 | ## Introduction 4 | 5 | ``` 6 | @inproceedings{lu2019grid, 7 | title={Grid r-cnn}, 8 | author={Lu, Xin and Li, Buyu and Yue, Yuxin and Li, Quanquan and Yan, Junjie}, 9 | booktitle={Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition}, 10 | year={2019} 11 | } 12 | 13 | @article{lu2019grid, 14 | title={Grid R-CNN Plus: Faster and Better}, 15 | author={Lu, Xin and Li, Buyu and Yue, Yuxin and Li, Quanquan and Yan, Junjie}, 16 | journal={arXiv preprint arXiv:1906.05688}, 17 | year={2019} 18 | } 19 | ``` 20 | 21 | ## Results and Models 22 | 23 | | Backbone | Lr schd | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | Download | 24 | |:-----------:|:-------:|:--------:|:-------------------:|:--------------:|:------:|:--------:| 25 | | R-50 | 2x | 4.8 | 1.172 | 10.9 | 40.3 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/grid_rcnn/grid_rcnn_gn_head_r50_fpn_2x_20190619-5b29cf9d.pth) | 26 | | R-101 | 2x | 6.7 | 1.214 | 10.0 | 41.7 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/grid_rcnn/grid_rcnn_gn_head_r101_fpn_2x_20190619-a4b61645.pth) | 27 | | X-101-32x4d | 2x | 8.0 | 1.335 | 8.5 | 43.0 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/grid_rcnn/grid_rcnn_gn_head_x101_32x4d_fpn_2x_20190619-0bbfd87a.pth) | 28 | | X-101-64x4d | 2x | 10.9 | 1.753 | 6.4 | 43.1 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/grid_rcnn/grid_rcnn_gn_head_x101_64x4d_fpn_2x_20190619-8f4e20bb.pth) | 29 | 30 | **Notes:** 31 | - All models are trained with 8 GPUs instead of 32 GPUs in the original paper. 32 | - The warming up lasts for 1 epoch and `2x` here indicates 25 epochs. 33 | -------------------------------------------------------------------------------- /configs/libra_rcnn/README.md: -------------------------------------------------------------------------------- 1 | # Libra R-CNN: Towards Balanced Learning for Object Detection 2 | 3 | ## Introduction 4 | 5 | We provide config files to reproduce the results in the CVPR 2019 paper [Libra R-CNN](https://arxiv.org/pdf/1904.02701.pdf). 6 | 7 | ``` 8 | @inproceedings{pang2019libra, 9 | title={Libra R-CNN: Towards Balanced Learning for Object Detection}, 10 | author={Pang, Jiangmiao and Chen, Kai and Shi, Jianping and Feng, Huajun and Ouyang, Wanli and Dahua Lin}, 11 | booktitle={IEEE Conference on Computer Vision and Pattern Recognition}, 12 | year={2019} 13 | } 14 | ``` 15 | 16 | ## Results and models 17 | 18 | The results on COCO 2017val are shown in the below table. (results on test-dev are usually slightly higher than val) 19 | 20 | | Architecture | Backbone | Style | Lr schd | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | Download | 21 | |:---------:|:-------:|:-------:|:--------:|:-------------------:|:--------------:|:------:|:-------:|:--------:| 22 | | Faster R-CNN | R-50-FPN | pytorch | 1x | 4.2 | 0.375 | 12.0 | 38.5 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/libra_rcnn/libra_faster_rcnn_r50_fpn_1x_20190610-bf0ea559.pth) | 23 | | Fast R-CNN | R-50-FPN | pytorch | 1x | 3.7 | 0.272 | 16.3 | 38.5 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/libra_rcnn/libra_fast_rcnn_r50_fpn_1x_20190525-a43f88b5.pth) | 24 | | Faster R-CNN | R-101-FPN | pytorch | 1x | 6.0 | 0.495 | 10.4 | 40.3 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/libra_rcnn/libra_faster_rcnn_r101_fpn_1x_20190525-94e94051.pth) | 25 | | Faster R-CNN | X-101-64x4d-FPN | pytorch | 1x | 10.1 | 1.050 | 6.8 | 42.7 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/libra_rcnn/libra_faster_rcnn_x101_64x4d_fpn_1x_20190525-359c134a.pth) | 26 | | RetinaNet | R-50-FPN | pytorch | 1x | 3.7 | 0.328 | 11.8 | 37.7 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/libra_rcnn/libra_retinanet_r50_fpn_1x_20190525-ead2a6bb.pth) | 27 | -------------------------------------------------------------------------------- /configs/ms_rcnn/README.md: -------------------------------------------------------------------------------- 1 | # Mask Scoring R-CNN 2 | 3 | ## Introduction 4 | 5 | ``` 6 | @inproceedings{huang2019msrcnn, 7 | title={Mask Scoring R-CNN}, 8 | author={Zhaojin Huang and Lichao Huang and Yongchao Gong and Chang Huang and Xinggang Wang}, 9 | booktitle={IEEE Conference on Computer Vision and Pattern Recognition}, 10 | year={2019}, 11 | } 12 | ``` 13 | 14 | ## Results and Models 15 | 16 | | Backbone | style | Lr schd | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | mask AP | Download | 17 | |:-------------:|:----------:|:-------:|:--------:|:-------------------:|:--------------:|:------:|:-------:|:--------:| 18 | | R-50-FPN | caffe | 1x | 4.3 | 0.537 | 10.1 | 37.4 | 35.5 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/ms-rcnn/ms_rcnn_r50_caffe_fpn_1x_20190624-619934b5.pth) | 19 | | R-50-FPN | caffe | 2x | - | - | - | 38.2 | 35.9 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/ms-rcnn/ms_rcnn_r50_caffe_fpn_2x_20190525-a07be31e.pth) | 20 | | R-101-FPN | caffe | 1x | 6.2 | 0.682 | 9.1 | 39.8 | 37.2 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/ms-rcnn/ms_rcnn_r101_caffe_fpn_1x_20190624-677a5548.pth) | 21 | | R-101-FPN | caffe | 2x | - | - | - | 40.7 | 37.8 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/ms-rcnn/ms_rcnn_r101_caffe_fpn_2x_20190525-4aee1528.pth) | 22 | | R-X101-32x4d | pytorch | 2x | 7.6 | 0.844 | 8.0 | 41.7 | 38.5 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/ms-rcnn/ms_rcnn_x101_32x4d_fpn_2x_20190628-ab454d07.pth) | 23 | | R-X101-64x4d | pytorch | 1x | 10.5 | 1.214 | 6.4 | 42.0 | 39.1 | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/ms-rcnn/ms_rcnn_x101_64x4d_fpn_1x_20190628-dec32bda.pth) | 24 | | R-X101-64x4d | pytorch | 2x | - | - | - | 42.2 | 38.9 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/ms-rcnn/ms_rcnn_x101_64x4d_fpn_2x_20190525-c044c25a.pth) | 25 | -------------------------------------------------------------------------------- /configs/pascal_voc/README.md: -------------------------------------------------------------------------------- 1 | ### SSD 2 | 3 | | Backbone | Size | Style | Lr schd | Mem (GB) | Train time (s/iter) | Inf time (fps) | box AP | Download | 4 | | :------: | :---: | :---: | :-----: | :------: | :-----------------: | :------------: | :----: | :------------------------------------------------------------------------------------------------------------------------------: | 5 | | VGG16 | 300 | caffe | 240e | 2.5 | 0.159 | 35.7 / 53.6 | 77.5 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/ssd300_voc_vgg16_caffe_240e_20190501-7160d09a.pth) | 6 | | VGG16 | 512 | caffe | 240e | 4.3 | 0.214 | 27.5 / 35.9 | 80.0 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/ssd512_voc_vgg16_caffe_240e_20190501-ff194be1.pth) | -------------------------------------------------------------------------------- /configs/reppoints/reppoints.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/csuhan/s2anet/e463c2bb17e161a6909e3d83410b49697dce770a/configs/reppoints/reppoints.png -------------------------------------------------------------------------------- /configs/rotated_iou/README.md: -------------------------------------------------------------------------------- 1 | ## Rotated IoU Loss 2 | 3 | ### Differentiable IoU between Polygons 4 | 5 | In general, IoU calculation for rotated boxes [1] is not differentiable as it requires [triangulation](https://en.wikipedia.org/wiki/Triangulation) to calculate the area of intesection polygons, like the [box_iou_rotated](mmdet/ops/box_iou_rotated) op. 6 | 7 | [2] proposes an IoU Loss for rotated 2D/3D objects, however, both implementation details and codes are not provided. 8 | [3] shows to calculate IoU between polygons in a pixel-wise manner. 9 | 10 | Recently, we find two implementations for differantiable IoU calculation, [4] and [5]. 11 | 12 | Here we use the picture in [4] to illustrate this method. 13 | 1. Find intesection points of two polygons. 14 | 2. Sorting these points to get the indices. 15 | 3. Using `torch.gather` to fetch the real values of points. 16 | 4. Calculating intesection areas by Shoelace_formula [6]. 17 | 18 | As the `torch.gather` is differentiable, the whole function is differentiable. 19 | 20 | 21 | 22 | 23 | ### IoU Loss for Oriented Object Detection 24 | 25 | We modify the code in [4] to our codebase and evaluate its performance on DOTA and HRSC2016. Specifically, we add an op called `box_iou_rotated_diff` lies in [here](mmdet/ops/box_iou_rotated_diff). 26 | 27 | The results show that optimizing bbox regression with IoU Loss can further boost the performance. Here we list the results in the table below. 28 | 29 | |Model |Data | Backbone | reg. loss | Rotate | Lr schd | box AP | 30 | |:-------------: |:-------------:| :-------------: | :--------: | :-----: | :-----: | :----: | 31 | |RetinaNet |HRSC2016 | R-50-FPN | smooth l1 | ✓ | 6x | 81.63 | 32 | |RetinaNet |HRSC2016 | R-50-FPN | IoU | ✓ | 6x | **82.74** | 33 | |CS2A-Net-2s |DOTA | R-50-FPN | smooth l1 | - | 1x | 73.83 | 34 | |CS2A-Net-2s |DOTA | R-50-FPN | IoU | - | 1x | **74.58** | 35 | 36 | 37 | **References** 38 | 39 | [1] Arbitrary-oriented scene text detection via rotation proposals 40 | 41 | [2] IoU Loss for 2D/3D Object Detection 42 | 43 | [3] PIoU Loss: Towards Accurate Oriented Object Detection in Complex Environments 44 | 45 | [4] https://github.com/lilanxiao/Rotated_IoU 46 | 47 | [5] https://github.com/maudzung/Complex-YOLOv4-Pytorch 48 | 49 | [6] https://en.wikipedia.org/wiki/Shoelace_formula -------------------------------------------------------------------------------- /configs/scratch/README.md: -------------------------------------------------------------------------------- 1 | # Rethinking ImageNet Pre-training 2 | 3 | ## Introduction 4 | 5 | ``` 6 | @article{he2018rethinking, 7 | title={Rethinking imagenet pre-training}, 8 | author={He, Kaiming and Girshick, Ross and Doll{\'a}r, Piotr}, 9 | journal={arXiv preprint arXiv:1811.08883}, 10 | year={2018} 11 | } 12 | ``` 13 | 14 | ## Results and Models 15 | 16 | | Model | Backbone | Style | Lr schd | box AP | mask AP | Download | 17 | |:------------:|:---------:|:-------:|:-------:|:------:|:-------:|:--------:| 18 | | Faster R-CNN | R-50-FPN | pytorch | 6x | 40.1 | - | [model](https://open-mmlab.s3.ap-northeast-2.amazonaws.com/mmdetection/models/scratch/scratch_faster_rcnn_r50_fpn_gn_6x_20190515-ff554978.pth) | 19 | | Mask R-CNN | R-50-FPN | pytorch | 6x | 41.0 | 37.4 | [model](https://s3.ap-northeast-2.amazonaws.com/open-mmlab/mmdetection/models/scratch/scratch_mask_rcnn_r50_fpn_gn_6x_20190515-96743f5e.pth) | 20 | 21 | Note: 22 | - The above models are trained with 16 GPUs. -------------------------------------------------------------------------------- /configs/wider_face/README.md: -------------------------------------------------------------------------------- 1 | ## WIDER Face Dataset 2 | 3 | To use the WIDER Face dataset you need to download it 4 | and extract to the `data/WIDERFace` folder. Annotation in the VOC format 5 | can be found in this [repo](https://github.com/sovrasov/wider-face-pascal-voc-annotations.git). 6 | You should move the annotation files from `WIDER_train_annotations` and `WIDER_val_annotations` folders 7 | to the `Annotation` folders inside the corresponding directories `WIDER_train` and `WIDER_val`. 8 | Also annotation lists `val.txt` and `train.txt` should be copied to `data/WIDERFace` from `WIDER_train_annotations` and `WIDER_val_annotations`. 9 | The directory should be like this: 10 | 11 | ``` 12 | mmdetection 13 | ├── mmdet 14 | ├── tools 15 | ├── configs 16 | ├── data 17 | │ ├── WIDERFace 18 | │ │ ├── WIDER_train 19 | │ | │ ├──0--Parade 20 | │ | │ ├── ... 21 | │ | │ ├── Annotations 22 | │ │ ├── WIDER_val 23 | │ | │ ├──0--Parade 24 | │ | │ ├── ... 25 | │ | │ ├── Annotations 26 | │ │ ├── val.txt 27 | │ │ ├── train.txt 28 | 29 | ``` 30 | 31 | After that you can train the SSD300 on WIDER by launching training with the `ssd300_wider_face.py` config or 32 | create your own config based on the presented one. 33 | -------------------------------------------------------------------------------- /demo/cascade_s2anet.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/csuhan/s2anet/e463c2bb17e161a6909e3d83410b49697dce770a/demo/cascade_s2anet.png -------------------------------------------------------------------------------- /demo/corruptions_sev_3.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/csuhan/s2anet/e463c2bb17e161a6909e3d83410b49697dce770a/demo/corruptions_sev_3.png -------------------------------------------------------------------------------- /demo/data_pipeline.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/csuhan/s2anet/e463c2bb17e161a6909e3d83410b49697dce770a/demo/data_pipeline.png -------------------------------------------------------------------------------- /demo/demo.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/csuhan/s2anet/e463c2bb17e161a6909e3d83410b49697dce770a/demo/demo.jpg -------------------------------------------------------------------------------- /demo/network.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/csuhan/s2anet/e463c2bb17e161a6909e3d83410b49697dce770a/demo/network.png -------------------------------------------------------------------------------- /demo/webcam_demo.py: -------------------------------------------------------------------------------- 1 | import argparse 2 | 3 | import cv2 4 | import torch 5 | 6 | from mmdet.apis import inference_detector, init_detector, show_result 7 | 8 | 9 | def parse_args(): 10 | parser = argparse.ArgumentParser(description='MMDetection webcam demo') 11 | parser.add_argument('config', help='test config file path') 12 | parser.add_argument('checkpoint', help='checkpoint file') 13 | parser.add_argument('--device', type=int, default=0, help='CUDA device id') 14 | parser.add_argument( 15 | '--camera-id', type=int, default=0, help='camera device id') 16 | parser.add_argument( 17 | '--score-thr', type=float, default=0.5, help='bbox score threshold') 18 | args = parser.parse_args() 19 | return args 20 | 21 | 22 | def main(): 23 | args = parse_args() 24 | 25 | model = init_detector( 26 | args.config, args.checkpoint, device=torch.device('cuda', args.device)) 27 | 28 | camera = cv2.VideoCapture(args.camera_id) 29 | 30 | print('Press "Esc", "q" or "Q" to exit.') 31 | while True: 32 | ret_val, img = camera.read() 33 | result = inference_detector(model, img) 34 | 35 | ch = cv2.waitKey(1) 36 | if ch == 27 or ch == ord('q') or ch == ord('Q'): 37 | break 38 | 39 | show_result( 40 | img, result, model.CLASSES, score_thr=args.score_thr, wait_time=1) 41 | 42 | 43 | if __name__ == '__main__': 44 | main() 45 | -------------------------------------------------------------------------------- /docker/Dockerfile: -------------------------------------------------------------------------------- 1 | ARG PYTORCH="1.3" 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 | RUN apt-get update && apt-get install -y libglib2.0-0 libsm6 libxrender-dev libxext6 swig\ 12 | && apt-get clean \ 13 | && rm -rf /var/lib/apt/lists/* 14 | 15 | # Install mmdetection 16 | RUN conda install cython -y && conda clean --all 17 | RUN git clone https://github.com/csuhan/s2anet.git /s2anet 18 | WORKDIR /s2anet 19 | RUN pip install --no-cache-dir -e . 20 | 21 | # install DOTA_devkit 22 | RUN cd DOTA_devkit/polyiou && swig -c++ -python csrc/polyiou.i && python setup.py build_ext --inplace -------------------------------------------------------------------------------- /docs/CASCADE_S2ANET.md: -------------------------------------------------------------------------------- 1 | ## Cascade S2A-Net 2 | 3 | Cascade S2A-Net (CS2A-Net) can be regarded as a generalized S2A-Net. 4 | It consists of several detection heads, and each head contains an Alignment Convolution and two subnetworks, _.i.e_, _cls_ and _reg_. 5 | The overall archetecture of CS2A-Net is shown below. 6 | 7 | ![](../demo/cascade_s2anet.png) 8 | 9 | The difference between CS2A-Net and S2A-Net is: 10 | * CS2A-Net head aligns features first by an Alignment Convolution Layer (ACL), and produce _cls_score_ and _bbox_pred_ later. 11 | While the FAM in S2A-Net does clssification and regression first, then produces aligned features. 12 | 13 | * CS2A-Net aligns features in each stage, while feature alignment in S2A-Net is only appeared at the end of FAM. 14 | So even the one stage version of CS2A-Net, _.i.e_, CS2A-Net-1s can produce aligned features. 15 | -------------------------------------------------------------------------------- /mmdet/__init__.py: -------------------------------------------------------------------------------- 1 | from .version import __version__, short_version 2 | 3 | __all__ = ['__version__', 'short_version'] 4 | -------------------------------------------------------------------------------- /mmdet/apis/__init__.py: -------------------------------------------------------------------------------- 1 | from .env import get_root_logger, init_dist, set_random_seed 2 | from .inference import (inference_detector, init_detector, show_result, 3 | show_result_pyplot) 4 | from .train import train_detector 5 | 6 | __all__ = [ 7 | 'init_dist', 'get_root_logger', 'set_random_seed', 'train_detector', 8 | 'init_detector', 'inference_detector', 'show_result', 'show_result_pyplot', 9 | ] 10 | -------------------------------------------------------------------------------- /mmdet/apis/env.py: -------------------------------------------------------------------------------- 1 | import logging 2 | import os 3 | import random 4 | import subprocess 5 | 6 | import numpy as np 7 | import torch 8 | import torch.distributed as dist 9 | import torch.multiprocessing as mp 10 | from mmcv.runner import get_dist_info 11 | 12 | 13 | def init_dist(launcher, backend='nccl', **kwargs): 14 | if mp.get_start_method(allow_none=True) is None: 15 | mp.set_start_method('spawn') 16 | if launcher == 'pytorch': 17 | _init_dist_pytorch(backend, **kwargs) 18 | elif launcher == 'mpi': 19 | _init_dist_mpi(backend, **kwargs) 20 | elif launcher == 'slurm': 21 | _init_dist_slurm(backend, **kwargs) 22 | else: 23 | raise ValueError('Invalid launcher type: {}'.format(launcher)) 24 | 25 | 26 | def _init_dist_pytorch(backend, **kwargs): 27 | # TODO: use local_rank instead of rank % num_gpus 28 | rank = int(os.environ['RANK']) 29 | num_gpus = torch.cuda.device_count() 30 | torch.cuda.set_device(rank % num_gpus) 31 | dist.init_process_group(backend=backend, **kwargs) 32 | 33 | 34 | def _init_dist_mpi(backend, **kwargs): 35 | raise NotImplementedError 36 | 37 | 38 | def _init_dist_slurm(backend, port=29500, **kwargs): 39 | proc_id = int(os.environ['SLURM_PROCID']) 40 | ntasks = int(os.environ['SLURM_NTASKS']) 41 | node_list = os.environ['SLURM_NODELIST'] 42 | num_gpus = torch.cuda.device_count() 43 | torch.cuda.set_device(proc_id % num_gpus) 44 | addr = subprocess.getoutput( 45 | 'scontrol show hostname {} | head -n1'.format(node_list)) 46 | os.environ['MASTER_PORT'] = str(port) 47 | os.environ['MASTER_ADDR'] = addr 48 | os.environ['WORLD_SIZE'] = str(ntasks) 49 | os.environ['RANK'] = str(proc_id) 50 | dist.init_process_group(backend=backend) 51 | 52 | 53 | def set_random_seed(seed): 54 | random.seed(seed) 55 | np.random.seed(seed) 56 | torch.manual_seed(seed) 57 | torch.cuda.manual_seed_all(seed) 58 | 59 | 60 | def get_root_logger(log_level=logging.INFO): 61 | logger = logging.getLogger() 62 | if not logger.hasHandlers(): 63 | logging.basicConfig( 64 | format='%(asctime)s - %(levelname)s - %(message)s', 65 | level=log_level) 66 | rank, _ = get_dist_info() 67 | if rank != 0: 68 | logger.setLevel('ERROR') 69 | return logger 70 | -------------------------------------------------------------------------------- /mmdet/core/__init__.py: -------------------------------------------------------------------------------- 1 | from .anchor import * # noqa: F401, F403 2 | from .bbox import * # noqa: F401, F403 3 | from .evaluation import * # noqa: F401, F403 4 | from .fp16 import * # noqa: F401, F403 5 | from .mask import * # noqa: F401, F403 6 | from .post_processing import * # noqa: F401, F403 7 | from .utils import * # noqa: F401, F403 8 | -------------------------------------------------------------------------------- /mmdet/core/anchor/__init__.py: -------------------------------------------------------------------------------- 1 | from .anchor_generator import AnchorGenerator 2 | from .anchor_generator_rotated import AnchorGeneratorRotated 3 | from .anchor_target import anchor_inside_flags, anchor_target, unmap, images_to_levels 4 | from .guided_anchor_target import ga_loc_target, ga_shape_target 5 | from .point_generator import PointGenerator 6 | from .point_target import point_target 7 | 8 | __all__ = [ 9 | 'AnchorGenerator', 'anchor_target', 'anchor_inside_flags', 'ga_loc_target', 10 | 'ga_shape_target', 'PointGenerator', 'point_target', 11 | 'unmap', 'images_to_levels', 'AnchorGeneratorRotated' 12 | ] 13 | -------------------------------------------------------------------------------- /mmdet/core/anchor/point_generator.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | 4 | class PointGenerator(object): 5 | 6 | def _meshgrid(self, x, y, row_major=True): 7 | xx = x.repeat(len(y)) 8 | yy = y.view(-1, 1).repeat(1, len(x)).view(-1) 9 | if row_major: 10 | return xx, yy 11 | else: 12 | return yy, xx 13 | 14 | def grid_points(self, featmap_size, stride=16, device='cuda'): 15 | feat_h, feat_w = featmap_size 16 | shift_x = torch.arange(0., feat_w, device=device) * stride 17 | shift_y = torch.arange(0., feat_h, device=device) * stride 18 | shift_xx, shift_yy = self._meshgrid(shift_x, shift_y) 19 | stride = shift_x.new_full((shift_xx.shape[0], ), stride) 20 | shifts = torch.stack([shift_xx, shift_yy, stride], dim=-1) 21 | all_points = shifts.to(device) 22 | return all_points 23 | 24 | def valid_flags(self, featmap_size, valid_size, device='cuda'): 25 | feat_h, feat_w = featmap_size 26 | valid_h, valid_w = valid_size 27 | assert valid_h <= feat_h and valid_w <= feat_w 28 | valid_x = torch.zeros(feat_w, dtype=torch.uint8, device=device) 29 | valid_y = torch.zeros(feat_h, dtype=torch.uint8, device=device) 30 | valid_x[:valid_w] = 1 31 | valid_y[:valid_h] = 1 32 | valid_xx, valid_yy = self._meshgrid(valid_x, valid_y) 33 | valid = valid_xx & valid_yy 34 | return valid 35 | -------------------------------------------------------------------------------- /mmdet/core/bbox/__init__.py: -------------------------------------------------------------------------------- 1 | from .assign_sampling import assign_and_sample 2 | from .assigners import AssignResult, BaseAssigner, MaxIoUAssigner 3 | from .bbox_target import bbox_target 4 | from .bbox_target_rotated import bbox_target_rotated 5 | from .builder import build_assigner, build_sampler, build_bbox_coder 6 | from .coder import DeltaXYWHBBoxCoder, DeltaXYWHABBoxCoder, PseudoBBoxCoder 7 | from .iou_calculators import bbox_overlaps, bbox_overlaps_rotated 8 | from .samplers import (BaseSampler, CombinedSampler, 9 | InstanceBalancedPosSampler, IoUBalancedNegSampler, 10 | PseudoSampler, RandomSampler, SamplingResult) 11 | from .transforms import (bbox2delta, bbox2result, bbox2roi, bbox_flip, 12 | bbox_mapping, bbox_mapping_back, delta2bbox, 13 | distance2bbox, roi2bbox) 14 | from .transforms_rotated import (norm_angle, 15 | poly_to_rotated_box_np, poly_to_rotated_box_single, poly_to_rotated_box, 16 | rotated_box_to_poly_np, rotated_box_to_poly_single, 17 | rotated_box_to_poly, rotated_box_to_bbox_np, rotated_box_to_bbox, 18 | bbox2result_rotated, bbox_flip_rotated, bbox_mapping_rotated, 19 | bbox_mapping_back_rotated, bbox_to_rotated_box, roi_to_rotated_box, rotated_box_to_roi, 20 | bbox2delta_rotated, delta2bbox_rotated) 21 | 22 | __all__ = [ 23 | 'BaseAssigner', 'MaxIoUAssigner', 'AssignResult', 24 | 'BaseSampler', 'PseudoSampler', 'RandomSampler', 25 | 'InstanceBalancedPosSampler', 'IoUBalancedNegSampler', 'CombinedSampler', 26 | 'SamplingResult', 'build_assigner', 'build_sampler', 'build_bbox_coder', 'assign_and_sample', 27 | 'bbox2delta', 'delta2bbox', 'bbox_flip', 'bbox_mapping', 28 | 'bbox_mapping_back', 'bbox2roi', 'roi2bbox', 'bbox2result', 29 | 'distance2bbox', 'bbox_target', 'bbox_flip_rotated', 'bbox2delta_rotated', 'delta2bbox_rotated', 30 | 'bbox_mapping_rotated', 'bbox_mapping_back_rotated', 'bbox2result_rotated', 31 | 'rotated_box_to_poly_np', 'poly_to_rotated_box_np', 'poly_to_rotated_box', 'rotated_box_to_poly', 32 | 'rotated_box_to_bbox_np', 'rotated_box_to_bbox', 'bbox_to_rotated_box', 'poly_to_rotated_box_single', 33 | 'rotated_box_to_poly_single', 'roi_to_rotated_box', 'rotated_box_to_roi', 'norm_angle', 34 | 'DeltaXYWHABBoxCoder', 'DeltaXYWHBBoxCoder', 'PseudoBBoxCoder', 35 | 'bbox_overlaps', 'bbox_overlaps_rotated', 'bbox_target_rotated' 36 | ] 37 | -------------------------------------------------------------------------------- /mmdet/core/bbox/assign_sampling.py: -------------------------------------------------------------------------------- 1 | from .builder import build_assigner, build_sampler 2 | 3 | 4 | def assign_and_sample(bboxes, gt_bboxes, gt_bboxes_ignore, gt_labels, cfg): 5 | bbox_assigner = build_assigner(cfg.assigner) 6 | bbox_sampler = build_sampler(cfg.sampler) 7 | assign_result = bbox_assigner.assign(bboxes, gt_bboxes, gt_bboxes_ignore, 8 | gt_labels) 9 | sampling_result = bbox_sampler.sample(assign_result, bboxes, gt_bboxes, 10 | gt_labels) 11 | return assign_result, sampling_result 12 | -------------------------------------------------------------------------------- /mmdet/core/bbox/assigners/__init__.py: -------------------------------------------------------------------------------- 1 | from .approx_max_iou_assigner import ApproxMaxIoUAssigner 2 | from .assign_result import AssignResult 3 | from .base_assigner import BaseAssigner 4 | from .max_iou_assigner import MaxIoUAssigner 5 | from .point_assigner import PointAssigner 6 | 7 | __all__ = [ 8 | 'BaseAssigner', 'MaxIoUAssigner', 'ApproxMaxIoUAssigner', 'AssignResult', 9 | 'PointAssigner' 10 | ] 11 | -------------------------------------------------------------------------------- /mmdet/core/bbox/assigners/assign_result.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | 4 | class AssignResult(object): 5 | 6 | def __init__(self, num_gts, gt_inds, max_overlaps, labels=None): 7 | self.num_gts = num_gts 8 | self.gt_inds = gt_inds 9 | self.max_overlaps = max_overlaps 10 | self.labels = labels 11 | 12 | def add_gt_(self, gt_labels): 13 | self_inds = torch.arange( 14 | 1, len(gt_labels) + 1, dtype=torch.long, device=gt_labels.device) 15 | self.gt_inds = torch.cat([self_inds, self.gt_inds]) 16 | self.max_overlaps = torch.cat( 17 | [self.max_overlaps.new_ones(self.num_gts), self.max_overlaps]) 18 | if self.labels is not None: 19 | self.labels = torch.cat([gt_labels, self.labels]) 20 | -------------------------------------------------------------------------------- /mmdet/core/bbox/assigners/base_assigner.py: -------------------------------------------------------------------------------- 1 | from abc import ABCMeta, abstractmethod 2 | 3 | 4 | class BaseAssigner(metaclass=ABCMeta): 5 | 6 | @abstractmethod 7 | def assign(self, bboxes, gt_bboxes, gt_bboxes_ignore=None, gt_labels=None): 8 | pass 9 | -------------------------------------------------------------------------------- /mmdet/core/bbox/bbox_target.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | from .transforms import bbox2delta 4 | from ..utils import multi_apply 5 | 6 | 7 | def bbox_target(pos_bboxes_list, 8 | neg_bboxes_list, 9 | pos_gt_bboxes_list, 10 | pos_gt_labels_list, 11 | cfg, 12 | reg_classes=1, 13 | target_means=[.0, .0, .0, .0], 14 | target_stds=[1.0, 1.0, 1.0, 1.0], 15 | concat=True): 16 | labels, label_weights, bbox_targets, bbox_weights = multi_apply( 17 | bbox_target_single, 18 | pos_bboxes_list, 19 | neg_bboxes_list, 20 | pos_gt_bboxes_list, 21 | pos_gt_labels_list, 22 | cfg=cfg, 23 | reg_classes=reg_classes, 24 | target_means=target_means, 25 | target_stds=target_stds) 26 | 27 | if concat: 28 | labels = torch.cat(labels, 0) 29 | label_weights = torch.cat(label_weights, 0) 30 | bbox_targets = torch.cat(bbox_targets, 0) 31 | bbox_weights = torch.cat(bbox_weights, 0) 32 | return labels, label_weights, bbox_targets, bbox_weights 33 | 34 | 35 | def bbox_target_single(pos_bboxes, 36 | neg_bboxes, 37 | pos_gt_bboxes, 38 | pos_gt_labels, 39 | cfg, 40 | reg_classes=1, 41 | target_means=[.0, .0, .0, .0], 42 | target_stds=[1.0, 1.0, 1.0, 1.0]): 43 | num_pos = pos_bboxes.size(0) 44 | num_neg = neg_bboxes.size(0) 45 | num_samples = num_pos + num_neg 46 | labels = pos_bboxes.new_zeros(num_samples, dtype=torch.long) 47 | label_weights = pos_bboxes.new_zeros(num_samples) 48 | bbox_targets = pos_bboxes.new_zeros(num_samples, 4) 49 | bbox_weights = pos_bboxes.new_zeros(num_samples, 4) 50 | if num_pos > 0: 51 | labels[:num_pos] = pos_gt_labels 52 | pos_weight = 1.0 if cfg.pos_weight <= 0 else cfg.pos_weight 53 | label_weights[:num_pos] = pos_weight 54 | pos_bbox_targets = bbox2delta(pos_bboxes, pos_gt_bboxes, target_means, 55 | target_stds) 56 | bbox_targets[:num_pos, :] = pos_bbox_targets 57 | bbox_weights[:num_pos, :] = 1 58 | if num_neg > 0: 59 | label_weights[-num_neg:] = 1.0 60 | return labels, label_weights, bbox_targets, bbox_weights 61 | 62 | 63 | def expand_target(bbox_targets, bbox_weights, labels, num_classes): 64 | bbox_targets_expand = bbox_targets.new_zeros( 65 | (bbox_targets.size(0), 4 * num_classes)) 66 | bbox_weights_expand = bbox_weights.new_zeros( 67 | (bbox_weights.size(0), 4 * num_classes)) 68 | for i in torch.nonzero(labels > 0).squeeze(-1): 69 | start, end = labels[i] * 4, (labels[i] + 1) * 4 70 | bbox_targets_expand[i, start:end] = bbox_targets[i, :] 71 | bbox_weights_expand[i, start:end] = bbox_weights[i, :] 72 | return bbox_targets_expand, bbox_weights_expand 73 | -------------------------------------------------------------------------------- /mmdet/core/bbox/bbox_target_rotated.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | from .transforms_rotated import bbox2delta_rotated 4 | from ..utils import multi_apply 5 | 6 | 7 | def bbox_target_rotated(pos_bboxes_list, 8 | neg_bboxes_list, 9 | pos_gt_bboxes_list, 10 | pos_gt_labels_list, 11 | cfg, 12 | reg_classes=1, 13 | target_means=[.0, .0, .0, .0, .0], 14 | target_stds=[1.0, 1.0, 1.0, 1.0, 1.0], 15 | concat=True): 16 | labels, label_weights, bbox_targets, bbox_weights = multi_apply( 17 | bbox_target_single, 18 | pos_bboxes_list, 19 | neg_bboxes_list, 20 | pos_gt_bboxes_list, 21 | pos_gt_labels_list, 22 | cfg=cfg, 23 | reg_classes=reg_classes, 24 | target_means=target_means, 25 | target_stds=target_stds) 26 | 27 | if concat: 28 | labels = torch.cat(labels, 0) 29 | label_weights = torch.cat(label_weights, 0) 30 | bbox_targets = torch.cat(bbox_targets, 0) 31 | bbox_weights = torch.cat(bbox_weights, 0) 32 | return labels, label_weights, bbox_targets, bbox_weights 33 | 34 | 35 | def bbox_target_single(pos_bboxes, 36 | neg_bboxes, 37 | pos_gt_bboxes, 38 | pos_gt_labels, 39 | cfg, 40 | reg_classes=1, 41 | target_means=[.0, .0, .0, .0, .0], 42 | target_stds=[1.0, 1.0, 1.0, 1.0, 1.0]): 43 | num_pos = pos_bboxes.size(0) 44 | num_neg = neg_bboxes.size(0) 45 | num_samples = num_pos + num_neg 46 | labels = pos_bboxes.new_zeros(num_samples, dtype=torch.long) 47 | label_weights = pos_bboxes.new_zeros(num_samples) 48 | bbox_targets = pos_bboxes.new_zeros(num_samples, 5) 49 | bbox_weights = pos_bboxes.new_zeros(num_samples, 5) 50 | if num_pos > 0: 51 | labels[:num_pos] = pos_gt_labels 52 | pos_weight = 1.0 if cfg.pos_weight <= 0 else cfg.pos_weight 53 | label_weights[:num_pos] = pos_weight 54 | pos_bbox_targets = bbox2delta_rotated(pos_bboxes, pos_gt_bboxes, target_means, 55 | target_stds) 56 | bbox_targets[:num_pos, :] = pos_bbox_targets 57 | bbox_weights[:num_pos, :] = 1 58 | if num_neg > 0: 59 | label_weights[-num_neg:] = 1.0 60 | return labels, label_weights, bbox_targets, bbox_weights 61 | -------------------------------------------------------------------------------- /mmdet/core/bbox/builder.py: -------------------------------------------------------------------------------- 1 | from mmdet.utils import Registry, build_from_cfg 2 | 3 | BBOX_ASSIGNERS = Registry('bbox_assigner') 4 | BBOX_SAMPLERS = Registry('bbox_sampler') 5 | BBOX_CODERS = Registry('bbox_coder') 6 | 7 | 8 | def build_assigner(cfg, **default_args): 9 | """Builder of box assigner.""" 10 | return build_from_cfg(cfg, BBOX_ASSIGNERS, default_args) 11 | 12 | 13 | def build_sampler(cfg, **default_args): 14 | """Builder of box sampler.""" 15 | return build_from_cfg(cfg, BBOX_SAMPLERS, default_args) 16 | 17 | 18 | def build_bbox_coder(cfg, **default_args): 19 | """Builder of box coder.""" 20 | return build_from_cfg(cfg, BBOX_CODERS, default_args) 21 | -------------------------------------------------------------------------------- /mmdet/core/bbox/coder/__init__.py: -------------------------------------------------------------------------------- 1 | from .base_bbox_coder import BaseBBoxCoder 2 | 3 | from .delta_xywh_bbox_coder import DeltaXYWHBBoxCoder 4 | from .pseudo_bbox_coder import PseudoBBoxCoder 5 | 6 | 7 | from .delta_xywha_bbox_coder import DeltaXYWHABBoxCoder 8 | 9 | __all__ = [ 10 | 'BaseBBoxCoder', 'PseudoBBoxCoder', 'DeltaXYWHBBoxCoder', 'DeltaXYWHABBoxCoder' 11 | ] 12 | -------------------------------------------------------------------------------- /mmdet/core/bbox/coder/base_bbox_coder.py: -------------------------------------------------------------------------------- 1 | from abc import ABCMeta, abstractmethod 2 | 3 | 4 | class BaseBBoxCoder(metaclass=ABCMeta): 5 | """Base bounding box coder.""" 6 | 7 | def __init__(self, **kwargs): 8 | pass 9 | 10 | @abstractmethod 11 | def encode(self, bboxes, gt_bboxes): 12 | """Encode deltas between bboxes and ground truth boxes.""" 13 | pass 14 | 15 | @abstractmethod 16 | def decode(self, bboxes, bboxes_pred): 17 | """Decode the predicted bboxes according to prediction and base 18 | boxes.""" 19 | pass 20 | -------------------------------------------------------------------------------- /mmdet/core/bbox/coder/delta_xywha_bbox_coder.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | from .base_bbox_coder import BaseBBoxCoder 4 | from ..builder import BBOX_CODERS 5 | from ..transforms_rotated import delta2bbox_rotated, bbox2delta_rotated 6 | 7 | 8 | @BBOX_CODERS.register_module 9 | class DeltaXYWHABBoxCoder(BaseBBoxCoder): 10 | """Delta XYWHA BBox coder. 11 | 12 | Following the practice in `R-CNN `_, 13 | this coder encodes bbox (x,y,w,h,a) into delta (dx, dy, dw, dh,da) and 14 | decodes delta (dx, dy, dw, dh,da) back to original bbox (x, y, w, h, a). 15 | 16 | Args: 17 | target_means (Sequence[float]): Denormalizing means of target for 18 | delta coordinates 19 | target_stds (Sequence[float]): Denormalizing standard deviation of 20 | target for delta coordinates 21 | clip_border (bool, optional): Whether clip the objects outside the 22 | border of the image. Defaults to True. 23 | """ 24 | 25 | def __init__(self, 26 | target_means=(0., 0., 0., 0., 0.), 27 | target_stds=(1., 1., 1., 1., 1.), 28 | clip_border=True): 29 | super(BaseBBoxCoder, self).__init__() 30 | self.means = target_means 31 | self.stds = target_stds 32 | self.clip_border = clip_border 33 | 34 | def encode(self, bboxes, gt_bboxes): 35 | """Get box regression transformation deltas that can be used to 36 | transform the ``bboxes`` into the ``gt_bboxes``. 37 | 38 | Args: 39 | bboxes (torch.Tensor): Source boxes, e.g., object proposals. 40 | gt_bboxes (torch.Tensor): Target of the transformation, e.g., 41 | ground-truth boxes. 42 | 43 | Returns: 44 | torch.Tensor: Box transformation deltas 45 | """ 46 | 47 | assert bboxes.size(0) == gt_bboxes.size(0) 48 | assert bboxes.size(-1) == gt_bboxes.size(-1) == 5 49 | encoded_bboxes = bbox2delta_rotated(bboxes, gt_bboxes, self.means, self.stds) 50 | return encoded_bboxes 51 | 52 | def decode(self, 53 | bboxes, 54 | pred_bboxes, 55 | max_shape=None, 56 | wh_ratio_clip=16 / 1000): 57 | """Apply transformation `pred_bboxes` to `boxes`. 58 | 59 | Args: 60 | boxes (torch.Tensor): Basic boxes. 61 | pred_bboxes (torch.Tensor): Encoded boxes with shape 62 | max_shape (tuple[int], optional): Maximum shape of boxes. 63 | Defaults to None. 64 | wh_ratio_clip (float, optional): The allowed ratio between 65 | width and height. 66 | 67 | Returns: 68 | torch.Tensor: Decoded boxes. 69 | """ 70 | assert pred_bboxes.size(0) == bboxes.size(0) 71 | decoded_bboxes = delta2bbox_rotated(bboxes, pred_bboxes, self.means, self.stds, 72 | max_shape, wh_ratio_clip, self.clip_border) 73 | 74 | return decoded_bboxes 75 | -------------------------------------------------------------------------------- /mmdet/core/bbox/coder/pseudo_bbox_coder.py: -------------------------------------------------------------------------------- 1 | from ..builder import BBOX_CODERS 2 | from .base_bbox_coder import BaseBBoxCoder 3 | 4 | 5 | @BBOX_CODERS.register_module 6 | class PseudoBBoxCoder(BaseBBoxCoder): 7 | """Pseudo bounding box coder.""" 8 | 9 | def __init__(self, **kwargs): 10 | super(BaseBBoxCoder, self).__init__(**kwargs) 11 | 12 | def encode(self, bboxes, gt_bboxes): 13 | """torch.Tensor: return the given ``bboxes``""" 14 | return gt_bboxes 15 | 16 | def decode(self, bboxes, pred_bboxes): 17 | """torch.Tensor: return the given ``pred_bboxes``""" 18 | return pred_bboxes 19 | -------------------------------------------------------------------------------- /mmdet/core/bbox/iou_calculators/__init__.py: -------------------------------------------------------------------------------- 1 | from .builder import build_iou_calculator 2 | from .iou2d_calculator import BboxOverlaps2D, bbox_overlaps 3 | from .iou2d_calculator_rotated import BboxOverlaps2D_rotated, bbox_overlaps_rotated 4 | 5 | __all__ = ['build_iou_calculator', 'BboxOverlaps2D', 'bbox_overlaps', 'BboxOverlaps2D_rotated', 'bbox_overlaps_rotated'] 6 | -------------------------------------------------------------------------------- /mmdet/core/bbox/iou_calculators/builder.py: -------------------------------------------------------------------------------- 1 | from mmdet.utils import Registry, build_from_cfg 2 | 3 | IOU_CALCULATORS = Registry('IoU calculator') 4 | 5 | 6 | def build_iou_calculator(cfg, default_args=None): 7 | """Builder of IoU calculator.""" 8 | return build_from_cfg(cfg, IOU_CALCULATORS, default_args) 9 | -------------------------------------------------------------------------------- /mmdet/core/bbox/iou_calculators/iou2d_calculator_rotated.py: -------------------------------------------------------------------------------- 1 | from mmdet.ops.box_iou_rotated import box_iou_rotated 2 | from .builder import IOU_CALCULATORS 3 | 4 | 5 | @IOU_CALCULATORS.register_module 6 | class BboxOverlaps2D_rotated(object): 7 | """2D Overlaps (e.g. IoUs, GIoUs) Calculator.""" 8 | 9 | def __call__(self, bboxes1, bboxes2, mode='iou', is_aligned=False): 10 | """Calculate IoU between 2D bboxes. 11 | 12 | Args: 13 | bboxes1 (Tensor): bboxes have shape (m, 5) in 14 | format, or shape (m, 5) in format. 15 | bboxes2 (Tensor): bboxes have shape (m, 5) in 16 | format, or shape (m, 5) in format, or be 17 | empty. If ``is_aligned `` is ``True``, then m and n must be 18 | equal. 19 | mode (str): "iou" (intersection over union), "iof" (intersection 20 | over foreground), or "giou" (generalized intersection over 21 | union). 22 | is_aligned (bool, optional): If True, then m and n must be equal. 23 | Default False. 24 | 25 | Returns: 26 | Tensor: shape (m, n) if ``is_aligned `` is False else shape (m,) 27 | """ 28 | assert bboxes1.size(-1) in [0, 5, 6] 29 | assert bboxes2.size(-1) in [0, 5, 6] 30 | if bboxes2.size(-1) == 6: 31 | bboxes2 = bboxes2[..., :5] 32 | if bboxes1.size(-1) == 6: 33 | bboxes1 = bboxes1[..., :5] 34 | return bbox_overlaps_rotated(bboxes1, bboxes2) 35 | 36 | def __repr__(self): 37 | """str: a string describing the module""" 38 | repr_str = self.__class__.__name__ + '()' 39 | return repr_str 40 | 41 | 42 | def bbox_overlaps_rotated(rboxes1, rboxes2): 43 | ious = box_iou_rotated(rboxes1.float(), rboxes2.float()) 44 | return ious 45 | -------------------------------------------------------------------------------- /mmdet/core/bbox/samplers/__init__.py: -------------------------------------------------------------------------------- 1 | from .base_sampler import BaseSampler 2 | from .combined_sampler import CombinedSampler 3 | from .instance_balanced_pos_sampler import InstanceBalancedPosSampler 4 | from .iou_balanced_neg_sampler import IoUBalancedNegSampler 5 | from .ohem_sampler import OHEMSampler 6 | from .pseudo_sampler import PseudoSampler 7 | from .random_sampler import RandomSampler 8 | from .random_sampler_rotated import RandomSamplerRotated 9 | from .sampling_result import SamplingResult 10 | 11 | __all__ = [ 12 | 'BaseSampler', 'PseudoSampler', 'RandomSampler', 13 | 'InstanceBalancedPosSampler', 'IoUBalancedNegSampler', 'CombinedSampler', 14 | 'OHEMSampler', 'SamplingResult', 'RandomSamplerRotated' 15 | ] 16 | -------------------------------------------------------------------------------- /mmdet/core/bbox/samplers/combined_sampler.py: -------------------------------------------------------------------------------- 1 | from ..builder import build_sampler 2 | from .base_sampler import BaseSampler 3 | 4 | 5 | class CombinedSampler(BaseSampler): 6 | 7 | def __init__(self, pos_sampler, neg_sampler, **kwargs): 8 | super(CombinedSampler, self).__init__(**kwargs) 9 | self.pos_sampler = build_sampler(pos_sampler, **kwargs) 10 | self.neg_sampler = build_sampler(neg_sampler, **kwargs) 11 | 12 | def _sample_pos(self, **kwargs): 13 | raise NotImplementedError 14 | 15 | def _sample_neg(self, **kwargs): 16 | raise NotImplementedError 17 | -------------------------------------------------------------------------------- /mmdet/core/bbox/samplers/instance_balanced_pos_sampler.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import torch 3 | 4 | from .random_sampler import RandomSampler 5 | 6 | 7 | class InstanceBalancedPosSampler(RandomSampler): 8 | 9 | def _sample_pos(self, assign_result, num_expected, **kwargs): 10 | pos_inds = torch.nonzero(assign_result.gt_inds > 0) 11 | if pos_inds.numel() != 0: 12 | pos_inds = pos_inds.squeeze(1) 13 | if pos_inds.numel() <= num_expected: 14 | return pos_inds 15 | else: 16 | unique_gt_inds = assign_result.gt_inds[pos_inds].unique() 17 | num_gts = len(unique_gt_inds) 18 | num_per_gt = int(round(num_expected / float(num_gts)) + 1) 19 | sampled_inds = [] 20 | for i in unique_gt_inds: 21 | inds = torch.nonzero(assign_result.gt_inds == i.item()) 22 | if inds.numel() != 0: 23 | inds = inds.squeeze(1) 24 | else: 25 | continue 26 | if len(inds) > num_per_gt: 27 | inds = self.random_choice(inds, num_per_gt) 28 | sampled_inds.append(inds) 29 | sampled_inds = torch.cat(sampled_inds) 30 | if len(sampled_inds) < num_expected: 31 | num_extra = num_expected - len(sampled_inds) 32 | extra_inds = np.array( 33 | list(set(pos_inds.cpu()) - set(sampled_inds.cpu()))) 34 | if len(extra_inds) > num_extra: 35 | extra_inds = self.random_choice(extra_inds, num_extra) 36 | extra_inds = torch.from_numpy(extra_inds).to( 37 | assign_result.gt_inds.device).long() 38 | sampled_inds = torch.cat([sampled_inds, extra_inds]) 39 | elif len(sampled_inds) > num_expected: 40 | sampled_inds = self.random_choice(sampled_inds, num_expected) 41 | return sampled_inds 42 | -------------------------------------------------------------------------------- /mmdet/core/bbox/samplers/pseudo_sampler.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | from .base_sampler import BaseSampler 4 | from .sampling_result import SamplingResult 5 | 6 | 7 | class PseudoSampler(BaseSampler): 8 | 9 | def __init__(self, **kwargs): 10 | pass 11 | 12 | def _sample_pos(self, **kwargs): 13 | raise NotImplementedError 14 | 15 | def _sample_neg(self, **kwargs): 16 | raise NotImplementedError 17 | 18 | def sample(self, assign_result, bboxes, gt_bboxes, **kwargs): 19 | pos_inds = torch.nonzero( 20 | assign_result.gt_inds > 0).squeeze(-1).unique() 21 | neg_inds = torch.nonzero( 22 | assign_result.gt_inds == 0).squeeze(-1).unique() 23 | gt_flags = bboxes.new_zeros(bboxes.shape[0], dtype=torch.uint8) 24 | sampling_result = SamplingResult(pos_inds, neg_inds, bboxes, gt_bboxes, 25 | assign_result, gt_flags) 26 | return sampling_result 27 | -------------------------------------------------------------------------------- /mmdet/core/bbox/samplers/random_sampler.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import torch 3 | from ..builder import BBOX_SAMPLERS 4 | 5 | from .base_sampler import BaseSampler 6 | 7 | @BBOX_SAMPLERS.register_module 8 | class RandomSampler(BaseSampler): 9 | 10 | def __init__(self, 11 | num, 12 | pos_fraction, 13 | neg_pos_ub=-1, 14 | add_gt_as_proposals=True, 15 | **kwargs): 16 | super(RandomSampler, self).__init__(num, pos_fraction, neg_pos_ub, 17 | add_gt_as_proposals) 18 | 19 | @staticmethod 20 | def random_choice(gallery, num): 21 | """Random select some elements from the gallery. 22 | 23 | It seems that Pytorch's implementation is slower than numpy so we use 24 | numpy to randperm the indices. 25 | """ 26 | assert len(gallery) >= num 27 | if isinstance(gallery, list): 28 | gallery = np.array(gallery) 29 | cands = np.arange(len(gallery)) 30 | np.random.shuffle(cands) 31 | rand_inds = cands[:num] 32 | if not isinstance(gallery, np.ndarray): 33 | rand_inds = torch.from_numpy(rand_inds).long().to(gallery.device) 34 | return gallery[rand_inds] 35 | 36 | def _sample_pos(self, assign_result, num_expected, **kwargs): 37 | """Randomly sample some positive samples.""" 38 | pos_inds = torch.nonzero(assign_result.gt_inds > 0) 39 | if pos_inds.numel() != 0: 40 | pos_inds = pos_inds.squeeze(1) 41 | if pos_inds.numel() <= num_expected: 42 | return pos_inds 43 | else: 44 | return self.random_choice(pos_inds, num_expected) 45 | 46 | def _sample_neg(self, assign_result, num_expected, **kwargs): 47 | """Randomly sample some negative samples.""" 48 | neg_inds = torch.nonzero(assign_result.gt_inds == 0) 49 | if neg_inds.numel() != 0: 50 | neg_inds = neg_inds.squeeze(1) 51 | if len(neg_inds) <= num_expected: 52 | return neg_inds 53 | else: 54 | return self.random_choice(neg_inds, num_expected) 55 | -------------------------------------------------------------------------------- /mmdet/core/bbox/samplers/random_sampler_rotated.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | from .random_sampler import RandomSampler 4 | from .sampling_result import SamplingResult 5 | from ..builder import BBOX_SAMPLERS 6 | 7 | 8 | @BBOX_SAMPLERS.register_module 9 | class RandomSamplerRotated(RandomSampler): 10 | 11 | def sample(self, 12 | assign_result, 13 | bboxes, 14 | gt_bboxes, 15 | gt_labels=None, 16 | **kwargs): 17 | 18 | gt_bboxes = gt_bboxes.float() 19 | bboxes = bboxes.float() 20 | 21 | if len(bboxes.shape) < 2: 22 | bboxes = bboxes[None, :] 23 | # this is the only difference between RandomSamplerRotated and RandomSampler 24 | bboxes = bboxes[:, :5] 25 | 26 | gt_flags = bboxes.new_zeros((bboxes.shape[0],), dtype=torch.uint8) 27 | if self.add_gt_as_proposals: 28 | bboxes = torch.cat([gt_bboxes, bboxes], dim=0) 29 | assign_result.add_gt_(gt_labels) 30 | gt_ones = bboxes.new_ones(gt_bboxes.shape[0], dtype=torch.uint8) 31 | gt_flags = torch.cat([gt_ones, gt_flags]) 32 | 33 | num_expected_pos = int(self.num * self.pos_fraction) 34 | pos_inds = self.pos_sampler._sample_pos( 35 | assign_result, num_expected_pos, bboxes=bboxes, **kwargs) 36 | # We found that sampled indices have duplicated items occasionally. 37 | # (may be a bug of PyTorch) 38 | pos_inds = pos_inds.unique() 39 | num_sampled_pos = pos_inds.numel() 40 | num_expected_neg = self.num - num_sampled_pos 41 | 42 | # print('Pos:{} Neg:{}'.format(num_sampled_pos,num_expected_neg)) 43 | 44 | if self.neg_pos_ub >= 0: 45 | _pos = max(1, num_sampled_pos) 46 | neg_upper_bound = int(self.neg_pos_ub * _pos) 47 | if num_expected_neg > neg_upper_bound: 48 | num_expected_neg = neg_upper_bound 49 | neg_inds = self.neg_sampler._sample_neg( 50 | assign_result, num_expected_neg, bboxes=bboxes, **kwargs) 51 | neg_inds = neg_inds.unique() 52 | 53 | return SamplingResult(pos_inds, neg_inds, bboxes, gt_bboxes, 54 | assign_result, gt_flags) 55 | -------------------------------------------------------------------------------- /mmdet/core/bbox/samplers/sampling_result.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | 4 | class SamplingResult(object): 5 | 6 | def __init__(self, pos_inds, neg_inds, bboxes, gt_bboxes, assign_result, 7 | gt_flags): 8 | self.pos_inds = pos_inds 9 | self.neg_inds = neg_inds 10 | self.pos_bboxes = bboxes[pos_inds] 11 | self.neg_bboxes = bboxes[neg_inds] 12 | self.pos_is_gt = gt_flags[pos_inds] 13 | 14 | self.num_gts = gt_bboxes.shape[0] 15 | self.pos_assigned_gt_inds = assign_result.gt_inds[pos_inds] - 1 16 | self.pos_gt_bboxes = gt_bboxes[self.pos_assigned_gt_inds, :] 17 | if assign_result.labels is not None: 18 | self.pos_gt_labels = assign_result.labels[pos_inds] 19 | else: 20 | self.pos_gt_labels = None 21 | 22 | @property 23 | def bboxes(self): 24 | return torch.cat([self.pos_bboxes, self.neg_bboxes]) 25 | -------------------------------------------------------------------------------- /mmdet/core/evaluation/__init__.py: -------------------------------------------------------------------------------- 1 | from .class_names import (coco_classes, dataset_aliases, get_classes, 2 | imagenet_det_classes, imagenet_vid_classes, 3 | voc_classes) 4 | from .coco_utils import coco_eval, fast_eval_recall, results2json 5 | from .eval_hooks import (CocoDistEvalmAPHook, CocoDistEvalRecallHook, 6 | DistEvalHook, DistEvalmAPHook) 7 | from .mean_ap import average_precision, eval_map, print_map_summary 8 | from .recall import (eval_recalls, plot_iou_recall, plot_num_recall, 9 | print_recall_summary) 10 | 11 | from .dota_utils import result2dota_task1,result2dota_task2 12 | 13 | __all__ = [ 14 | 'voc_classes', 'imagenet_det_classes', 'imagenet_vid_classes', 15 | 'coco_classes', 'dataset_aliases', 'get_classes', 'coco_eval', 16 | 'fast_eval_recall', 'results2json', 'DistEvalHook', 'DistEvalmAPHook', 17 | 'CocoDistEvalRecallHook', 'CocoDistEvalmAPHook', 'average_precision', 18 | 'eval_map', 'print_map_summary', 'eval_recalls', 'print_recall_summary', 19 | 'plot_num_recall', 'plot_iou_recall', 20 | 'result2dota_task1','result2dota_task2' 21 | ] 22 | -------------------------------------------------------------------------------- /mmdet/core/evaluation/bbox_overlaps.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | 3 | 4 | def bbox_overlaps(bboxes1, bboxes2, mode='iou'): 5 | """Calculate the ious between each bbox of bboxes1 and bboxes2. 6 | 7 | Args: 8 | bboxes1(ndarray): shape (n, 4) 9 | bboxes2(ndarray): shape (k, 4) 10 | mode(str): iou (intersection over union) or iof (intersection 11 | over foreground) 12 | 13 | Returns: 14 | ious(ndarray): shape (n, k) 15 | """ 16 | 17 | assert mode in ['iou', 'iof'] 18 | 19 | bboxes1 = bboxes1.astype(np.float32) 20 | bboxes2 = bboxes2.astype(np.float32) 21 | rows = bboxes1.shape[0] 22 | cols = bboxes2.shape[0] 23 | ious = np.zeros((rows, cols), dtype=np.float32) 24 | if rows * cols == 0: 25 | return ious 26 | exchange = False 27 | if bboxes1.shape[0] > bboxes2.shape[0]: 28 | bboxes1, bboxes2 = bboxes2, bboxes1 29 | ious = np.zeros((cols, rows), dtype=np.float32) 30 | exchange = True 31 | area1 = (bboxes1[:, 2] - bboxes1[:, 0] + 1) * ( 32 | bboxes1[:, 3] - bboxes1[:, 1] + 1) 33 | area2 = (bboxes2[:, 2] - bboxes2[:, 0] + 1) * ( 34 | bboxes2[:, 3] - bboxes2[:, 1] + 1) 35 | for i in range(bboxes1.shape[0]): 36 | x_start = np.maximum(bboxes1[i, 0], bboxes2[:, 0]) 37 | y_start = np.maximum(bboxes1[i, 1], bboxes2[:, 1]) 38 | x_end = np.minimum(bboxes1[i, 2], bboxes2[:, 2]) 39 | y_end = np.minimum(bboxes1[i, 3], bboxes2[:, 3]) 40 | overlap = np.maximum(x_end - x_start + 1, 0) * np.maximum( 41 | y_end - y_start + 1, 0) 42 | if mode == 'iou': 43 | union = area1[i] + area2 - overlap 44 | else: 45 | union = area1[i] if not exchange else area2 46 | ious[i, :] = overlap / union 47 | if exchange: 48 | ious = ious.T 49 | return ious 50 | -------------------------------------------------------------------------------- /mmdet/core/evaluation/dota_utils.py: -------------------------------------------------------------------------------- 1 | import os 2 | import os.path as osp 3 | 4 | from ..bbox import rotated_box_to_poly_single 5 | 6 | 7 | def result2dota_task1(results, dst_path, dataset): 8 | CLASSES = dataset.CLASSES 9 | img_names = dataset.img_names 10 | assert len(results) == len( 11 | img_names), 'length of results must equal with length of img_names' 12 | if not osp.exists(dst_path): 13 | os.mkdir(dst_path) 14 | for classname in CLASSES: 15 | f_out = open(osp.join(dst_path, 'Task1_'+classname+'.txt'), 'w') 16 | print('Task1_'+classname+'.txt') 17 | # per result represent one image 18 | for img_id, result in enumerate(results): 19 | for class_id, bboxes in enumerate(result): 20 | if CLASSES[class_id] != classname: 21 | continue 22 | if(bboxes.size != 0): 23 | for bbox in bboxes: 24 | score = bbox[5] 25 | bbox = rotated_box_to_poly_single(bbox[:5]) 26 | temp_txt = '{} {:.4f} {:.4f} {:.4f} {:.4f} {:.4f} {:.4f} {:.4f} {:.4f} {:.4f}\n'.format( 27 | osp.splitext(img_names[img_id])[0], score, bbox[0], bbox[1], bbox[2], bbox[3], bbox[4], bbox[5], bbox[6], bbox[7]) 28 | f_out.write(temp_txt) 29 | f_out.close() 30 | return True 31 | 32 | 33 | def result2dota_task2(results, dst_path, dataset): 34 | CLASSES = dataset.CLASSES 35 | img_names = dataset.img_names 36 | if not osp.exists(dst_path): 37 | os.mkdir(dst_path) 38 | for classname in CLASSES: 39 | f_out = open(osp.join(dst_path, 'Task2_'+classname+'.txt'), 'w') 40 | print('Task2_'+classname+'.txt') 41 | # per result represent one image 42 | for img_id, result in enumerate(results): 43 | filename = img_names[img_id] 44 | filename = osp.basename(filename) 45 | filename = osp.splitext(filename)[0] 46 | for class_id, bboxes in enumerate(result): 47 | if CLASSES[class_id] != classname: 48 | continue 49 | if(bboxes.size != 0): 50 | for bbox in bboxes: 51 | score = bbox[4] 52 | temp_txt = '{} {:.4f} {:.4f} {:.4f} {:.4f} {:.4f}\n'.format( 53 | filename, score, bbox[0], bbox[1], bbox[2], bbox[3]) 54 | f_out.write(temp_txt) 55 | f_out.close() 56 | return True 57 | -------------------------------------------------------------------------------- /mmdet/core/fp16/__init__.py: -------------------------------------------------------------------------------- 1 | from .decorators import auto_fp16, force_fp32 2 | from .hooks import Fp16OptimizerHook, wrap_fp16_model 3 | 4 | __all__ = ['auto_fp16', 'force_fp32', 'Fp16OptimizerHook', 'wrap_fp16_model'] 5 | -------------------------------------------------------------------------------- /mmdet/core/fp16/utils.py: -------------------------------------------------------------------------------- 1 | from collections import abc 2 | 3 | import numpy as np 4 | import torch 5 | 6 | 7 | def cast_tensor_type(inputs, src_type, dst_type): 8 | if isinstance(inputs, torch.Tensor): 9 | return inputs.to(dst_type) 10 | elif isinstance(inputs, str): 11 | return inputs 12 | elif isinstance(inputs, np.ndarray): 13 | return inputs 14 | elif isinstance(inputs, abc.Mapping): 15 | return type(inputs)({ 16 | k: cast_tensor_type(v, src_type, dst_type) 17 | for k, v in inputs.items() 18 | }) 19 | elif isinstance(inputs, abc.Iterable): 20 | return type(inputs)( 21 | cast_tensor_type(item, src_type, dst_type) for item in inputs) 22 | else: 23 | return inputs 24 | -------------------------------------------------------------------------------- /mmdet/core/mask/__init__.py: -------------------------------------------------------------------------------- 1 | from .mask_target import mask_target 2 | from .utils import split_combined_polys 3 | 4 | __all__ = ['split_combined_polys', 'mask_target'] 5 | -------------------------------------------------------------------------------- /mmdet/core/mask/mask_target.py: -------------------------------------------------------------------------------- 1 | import mmcv 2 | import numpy as np 3 | import torch 4 | from torch.nn.modules.utils import _pair 5 | 6 | 7 | def mask_target(pos_proposals_list, pos_assigned_gt_inds_list, gt_masks_list, 8 | cfg): 9 | cfg_list = [cfg for _ in range(len(pos_proposals_list))] 10 | mask_targets = map(mask_target_single, pos_proposals_list, 11 | pos_assigned_gt_inds_list, gt_masks_list, cfg_list) 12 | mask_targets = torch.cat(list(mask_targets)) 13 | return mask_targets 14 | 15 | 16 | def mask_target_single(pos_proposals, pos_assigned_gt_inds, gt_masks, cfg): 17 | mask_size = _pair(cfg.mask_size) 18 | num_pos = pos_proposals.size(0) 19 | mask_targets = [] 20 | if num_pos > 0: 21 | proposals_np = pos_proposals.cpu().numpy() 22 | pos_assigned_gt_inds = pos_assigned_gt_inds.cpu().numpy() 23 | for i in range(num_pos): 24 | gt_mask = gt_masks[pos_assigned_gt_inds[i]] 25 | bbox = proposals_np[i, :].astype(np.int32) 26 | x1, y1, x2, y2 = bbox 27 | w = np.maximum(x2 - x1 + 1, 1) 28 | h = np.maximum(y2 - y1 + 1, 1) 29 | # mask is uint8 both before and after resizing 30 | # mask_size (h, w) to (w, h) 31 | target = mmcv.imresize(gt_mask[y1:y1 + h, x1:x1 + w], 32 | mask_size[::-1]) 33 | mask_targets.append(target) 34 | mask_targets = torch.from_numpy(np.stack(mask_targets)).float().to( 35 | pos_proposals.device) 36 | else: 37 | mask_targets = pos_proposals.new_zeros((0, ) + mask_size) 38 | return mask_targets 39 | -------------------------------------------------------------------------------- /mmdet/core/mask/utils.py: -------------------------------------------------------------------------------- 1 | import mmcv 2 | 3 | 4 | def split_combined_polys(polys, poly_lens, polys_per_mask): 5 | """Split the combined 1-D polys into masks. 6 | 7 | A mask is represented as a list of polys, and a poly is represented as 8 | a 1-D array. In dataset, all masks are concatenated into a single 1-D 9 | tensor. Here we need to split the tensor into original representations. 10 | 11 | Args: 12 | polys (list): a list (length = image num) of 1-D tensors 13 | poly_lens (list): a list (length = image num) of poly length 14 | polys_per_mask (list): a list (length = image num) of poly number 15 | of each mask 16 | 17 | Returns: 18 | list: a list (length = image num) of list (length = mask num) of 19 | list (length = poly num) of numpy array 20 | """ 21 | mask_polys_list = [] 22 | for img_id in range(len(polys)): 23 | polys_single = polys[img_id] 24 | polys_lens_single = poly_lens[img_id].tolist() 25 | polys_per_mask_single = polys_per_mask[img_id].tolist() 26 | 27 | split_polys = mmcv.slice_list(polys_single, polys_lens_single) 28 | mask_polys = mmcv.slice_list(split_polys, polys_per_mask_single) 29 | mask_polys_list.append(mask_polys) 30 | return mask_polys_list 31 | -------------------------------------------------------------------------------- /mmdet/core/post_processing/__init__.py: -------------------------------------------------------------------------------- 1 | from .bbox_nms import multiclass_nms 2 | from .bbox_nms_rotated import multiclass_nms_rotated 3 | from .merge_augs import (merge_aug_bboxes, merge_aug_masks, 4 | merge_aug_proposals, merge_aug_scores) 5 | from .merge_augs_rotated import merge_aug_bboxes_rotated, merge_aug_proposals_rotated 6 | 7 | __all__ = [ 8 | 'multiclass_nms', 'merge_aug_proposals', 'merge_aug_bboxes', 9 | 'merge_aug_scores', 'merge_aug_masks', 10 | 'multiclass_nms_rotated', 'merge_aug_bboxes_rotated', 'merge_aug_proposals_rotated' 11 | ] 12 | -------------------------------------------------------------------------------- /mmdet/core/post_processing/bbox_nms.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | from mmdet.ops.nms import nms_wrapper 4 | 5 | 6 | def multiclass_nms(multi_bboxes, 7 | multi_scores, 8 | score_thr, 9 | nms_cfg, 10 | max_num=-1, 11 | score_factors=None): 12 | """NMS for multi-class bboxes. 13 | Args: 14 | multi_bboxes (Tensor): shape (n, #class*4) or (n, 4) 15 | multi_scores (Tensor): shape (n, #class), where the 0th column 16 | contains scores of the background class, but this will be ignored. 17 | score_thr (float): bbox threshold, bboxes with scores lower than it 18 | will not be considered. 19 | nms_thr (float): NMS IoU threshold 20 | max_num (int): if there are more than max_num bboxes after NMS, 21 | only top max_num will be kept. 22 | score_factors (Tensor): The factors multiplied to scores before 23 | applying NMS 24 | Returns: 25 | tuple: (bboxes, labels), tensors of shape (k, 5) and (k, 1). Labels 26 | are 0-based. 27 | """ 28 | num_classes = multi_scores.size(1) - 1 29 | # exclude background category 30 | if multi_bboxes.shape[1] > 4: 31 | bboxes = multi_bboxes.view(multi_scores.size(0), -1, 4)[:, 1:] 32 | else: 33 | bboxes = multi_bboxes[:, None].expand(-1, num_classes, 4) 34 | scores = multi_scores[:, 1:] 35 | 36 | # filter out boxes with low scores 37 | valid_mask = scores > score_thr 38 | bboxes = bboxes[valid_mask] 39 | if score_factors is not None: 40 | scores = scores * score_factors[:, None] 41 | scores = scores[valid_mask] 42 | labels = valid_mask.nonzero()[:, 1] 43 | 44 | if bboxes.numel() == 0: 45 | bboxes = multi_bboxes.new_zeros((0, 5)) 46 | labels = multi_bboxes.new_zeros((0, ), dtype=torch.long) 47 | return bboxes, labels 48 | 49 | # Modified from https://github.com/pytorch/vision/blob 50 | # /505cd6957711af790211896d32b40291bea1bc21/torchvision/ops/boxes.py#L39. 51 | # strategy: in order to perform NMS independently per class. 52 | # we add an offset to all the boxes. The offset is dependent 53 | # only on the class idx, and is large enough so that boxes 54 | # from different classes do not overlap 55 | max_coordinate = bboxes.max() 56 | offsets = labels.to(bboxes) * (max_coordinate + 1) 57 | bboxes_for_nms = bboxes + offsets[:, None] 58 | nms_cfg_ = nms_cfg.copy() 59 | nms_type = nms_cfg_.pop('type', 'nms') 60 | nms_op = getattr(nms_wrapper, nms_type) 61 | dets, keep = nms_op( 62 | torch.cat([bboxes_for_nms, scores[:, None]], 1), **nms_cfg_) 63 | bboxes = bboxes[keep] 64 | scores = dets[:, -1] # soft_nms will modify scores 65 | labels = labels[keep] 66 | 67 | if keep.size(0) > max_num: 68 | _, inds = scores.sort(descending=True) 69 | inds = inds[:max_num] 70 | bboxes = bboxes[inds] 71 | scores = scores[inds] 72 | labels = labels[inds] 73 | 74 | return torch.cat([bboxes, scores[:, None]], 1), labels -------------------------------------------------------------------------------- /mmdet/core/post_processing/bbox_nms_rotated.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | from mmdet.ops import ml_nms_rotated 4 | 5 | 6 | def multiclass_nms_rotated(multi_bboxes, 7 | multi_scores, 8 | score_thr, 9 | nms_cfg, 10 | max_num=-1, 11 | score_factors=None): 12 | """NMS for multi-class bboxes. 13 | Args: 14 | multi_bboxes (Tensor): shape (n, #class*5) or (n, 5) 15 | multi_scores (Tensor): shape (n, #class), where the 0th column 16 | contains scores of the background class, but this will be ignored. 17 | score_thr (float): bbox threshold, bboxes with scores lower than it 18 | will not be considered. 19 | nms_thr (float): NMS IoU threshold 20 | max_num (int): if there are more than max_num bboxes after NMS, 21 | only top max_num will be kept. 22 | score_factors (Tensor): The factors multiplied to scores before 23 | applying NMS 24 | Returns: 25 | tuple: (bboxes, labels), tensors of shape (k, 6) and (k, 1). Labels 26 | are 0-based. 27 | """ 28 | num_classes = multi_scores.size(1) - 1 29 | # exclude background category 30 | if multi_bboxes.shape[1] > 5: 31 | bboxes = multi_bboxes.view(multi_scores.size(0), -1, 5)[:, 1:] 32 | else: 33 | bboxes = multi_bboxes[:, None].expand(-1, num_classes, 5) 34 | scores = multi_scores[:, 1:] 35 | 36 | # filter out boxes with low scores 37 | valid_mask = scores > score_thr 38 | bboxes = bboxes[valid_mask] 39 | if score_factors is not None: 40 | scores = scores * score_factors[:, None] 41 | scores = scores[valid_mask] 42 | labels = valid_mask.nonzero()[:, 1] 43 | 44 | if bboxes.numel() == 0: 45 | bboxes = multi_bboxes.new_zeros((0, 6)) 46 | labels = multi_bboxes.new_zeros((0,), dtype=torch.long) 47 | return bboxes, labels 48 | nms_cfg_ = nms_cfg.copy() 49 | nms_type = nms_cfg_.pop('type', 'nms') 50 | iou_thr = nms_cfg_.pop('iou_thr', 0.1) 51 | labels = labels.to(bboxes) 52 | keep = ml_nms_rotated(bboxes, scores, labels, iou_thr) 53 | bboxes = bboxes[keep] 54 | scores = scores[keep] 55 | labels = labels[keep] 56 | 57 | if keep.size(0) > max_num: 58 | _, inds = scores.sort(descending=True) 59 | inds = inds[:max_num] 60 | bboxes = bboxes[inds] 61 | scores = scores[inds] 62 | labels = labels[inds] 63 | 64 | return torch.cat([bboxes, scores[:, None]], 1), labels 65 | -------------------------------------------------------------------------------- /mmdet/core/post_processing/merge_augs_rotated.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | from mmdet.ops import nms_rotated 4 | from ..bbox import bbox_mapping_back_rotated 5 | 6 | 7 | def merge_aug_proposals_rotated(aug_proposals, img_metas, rpn_test_cfg): 8 | """Merge augmented proposals (multiscale, flip, etc.) 9 | 10 | Args: 11 | aug_proposals (list[Tensor]): proposals from different testing 12 | schemes, shape (n, 6). Note that they are not rescaled to the 13 | original image size. 14 | 15 | img_metas (list[dict]): list of image info dict where each dict has: 16 | 'img_shape', 'scale_factor', 'flip', and my also contain 17 | 'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'. 18 | For details on the values of these keys see 19 | `mmdet/datasets/pipelines/formatting.py:Collect`. 20 | 21 | rpn_test_cfg (dict): rpn test config. 22 | 23 | Returns: 24 | Tensor: shape (n, 5), proposals corresponding to original image scale. 25 | """ 26 | recovered_proposals = [] 27 | for proposals, img_info in zip(aug_proposals, img_metas): 28 | img_shape = img_info['img_shape'] 29 | scale_factor = img_info['scale_factor'] 30 | flip = img_info['flip'] 31 | _proposals = proposals.clone() 32 | _proposals[:, :5] = bbox_mapping_back_rotated(_proposals[:, :5], img_shape, 33 | scale_factor, flip) 34 | recovered_proposals.append(_proposals) 35 | aug_proposals = torch.cat(recovered_proposals, dim=0) 36 | merged_proposals, _ = nms_rotated(aug_proposals, rpn_test_cfg.nms_thr) 37 | scores = merged_proposals[:, 5] 38 | _, order = scores.sort(0, descending=True) 39 | num = min(rpn_test_cfg.max_num, merged_proposals.shape[0]) 40 | order = order[:num] 41 | merged_proposals = merged_proposals[order, :] 42 | return merged_proposals 43 | 44 | 45 | def merge_aug_bboxes_rotated(aug_bboxes, aug_scores, img_metas, rcnn_test_cfg): 46 | """Merge augmented detection bboxes and scores. 47 | 48 | Args: 49 | aug_bboxes (list[Tensor]): shape (n, 5*#class) 50 | aug_scores (list[Tensor] or None): shape (n, #class) 51 | img_shapes (list[Tensor]): shape (3, ). 52 | rcnn_test_cfg (dict): rcnn test config. 53 | 54 | Returns: 55 | tuple: (bboxes, scores) 56 | """ 57 | recovered_bboxes = [] 58 | for bboxes, img_info in zip(aug_bboxes, img_metas): 59 | img_shape = img_info[0]['img_shape'] 60 | scale_factor = img_info[0]['scale_factor'] 61 | flip = img_info[0]['flip'] 62 | bboxes = bbox_mapping_back_rotated(bboxes, img_shape, scale_factor, flip) 63 | recovered_bboxes.append(bboxes) 64 | bboxes = torch.stack(recovered_bboxes).mean(dim=0) 65 | if aug_scores is None: 66 | return bboxes 67 | else: 68 | scores = torch.stack(aug_scores).mean(dim=0) 69 | return bboxes, scores 70 | -------------------------------------------------------------------------------- /mmdet/core/utils/__init__.py: -------------------------------------------------------------------------------- 1 | from .dist_utils import DistOptimizerHook, allreduce_grads 2 | from .misc import multi_apply, tensor2imgs, unmap 3 | 4 | __all__ = [ 5 | 'allreduce_grads', 'DistOptimizerHook', 'tensor2imgs', 'unmap', 6 | 'multi_apply' 7 | ] 8 | -------------------------------------------------------------------------------- /mmdet/core/utils/dist_utils.py: -------------------------------------------------------------------------------- 1 | from collections import OrderedDict 2 | 3 | import torch.distributed as dist 4 | from mmcv.runner import OptimizerHook 5 | from torch._utils import (_flatten_dense_tensors, _take_tensors, 6 | _unflatten_dense_tensors) 7 | 8 | 9 | def _allreduce_coalesced(tensors, world_size, bucket_size_mb=-1): 10 | if bucket_size_mb > 0: 11 | bucket_size_bytes = bucket_size_mb * 1024 * 1024 12 | buckets = _take_tensors(tensors, bucket_size_bytes) 13 | else: 14 | buckets = OrderedDict() 15 | for tensor in tensors: 16 | tp = tensor.type() 17 | if tp not in buckets: 18 | buckets[tp] = [] 19 | buckets[tp].append(tensor) 20 | buckets = buckets.values() 21 | 22 | for bucket in buckets: 23 | flat_tensors = _flatten_dense_tensors(bucket) 24 | dist.all_reduce(flat_tensors) 25 | flat_tensors.div_(world_size) 26 | for tensor, synced in zip( 27 | bucket, _unflatten_dense_tensors(flat_tensors, bucket)): 28 | tensor.copy_(synced) 29 | 30 | 31 | def allreduce_grads(params, coalesce=True, bucket_size_mb=-1): 32 | grads = [ 33 | param.grad.data for param in params 34 | if param.requires_grad and param.grad is not None 35 | ] 36 | world_size = dist.get_world_size() 37 | if coalesce: 38 | _allreduce_coalesced(grads, world_size, bucket_size_mb) 39 | else: 40 | for tensor in grads: 41 | dist.all_reduce(tensor.div_(world_size)) 42 | 43 | 44 | class DistOptimizerHook(OptimizerHook): 45 | 46 | def __init__(self, grad_clip=None, coalesce=True, bucket_size_mb=-1): 47 | self.grad_clip = grad_clip 48 | self.coalesce = coalesce 49 | self.bucket_size_mb = bucket_size_mb 50 | 51 | def after_train_iter(self, runner): 52 | runner.optimizer.zero_grad() 53 | runner.outputs['loss'].backward() 54 | allreduce_grads(runner.model.parameters(), self.coalesce, 55 | self.bucket_size_mb) 56 | if self.grad_clip is not None: 57 | self.clip_grads(runner.model.parameters()) 58 | runner.optimizer.step() 59 | -------------------------------------------------------------------------------- /mmdet/core/utils/misc.py: -------------------------------------------------------------------------------- 1 | from functools import partial 2 | 3 | import mmcv 4 | import numpy as np 5 | from six.moves import map, zip 6 | 7 | 8 | def tensor2imgs(tensor, mean=(0, 0, 0), std=(1, 1, 1), to_rgb=True): 9 | num_imgs = tensor.size(0) 10 | mean = np.array(mean, dtype=np.float32) 11 | std = np.array(std, dtype=np.float32) 12 | imgs = [] 13 | for img_id in range(num_imgs): 14 | img = tensor[img_id, ...].cpu().numpy().transpose(1, 2, 0) 15 | img = mmcv.imdenormalize( 16 | img, mean, std, to_bgr=to_rgb).astype(np.uint8) 17 | imgs.append(np.ascontiguousarray(img)) 18 | return imgs 19 | 20 | 21 | def multi_apply(func, *args, **kwargs): 22 | pfunc = partial(func, **kwargs) if kwargs else func 23 | map_results = map(pfunc, *args) 24 | return tuple(map(list, zip(*map_results))) 25 | 26 | 27 | def unmap(data, count, inds, fill=0): 28 | """ Unmap a subset of item (data) back to the original set of items (of 29 | size count) """ 30 | if data.dim() == 1: 31 | ret = data.new_full((count, ), fill) 32 | ret[inds] = data 33 | else: 34 | new_size = (count, ) + data.size()[1:] 35 | ret = data.new_full(new_size, fill) 36 | ret[inds, :] = data 37 | return ret 38 | -------------------------------------------------------------------------------- /mmdet/datasets/__init__.py: -------------------------------------------------------------------------------- 1 | from .builder import build_dataset 2 | from .cityscapes import CityscapesDataset 3 | from .coco import CocoDataset 4 | from .custom import CustomDataset 5 | from .dataset_wrappers import ConcatDataset, RepeatDataset 6 | from .loader import DistributedGroupSampler, GroupSampler, build_dataloader 7 | from .registry import DATASETS 8 | from .voc import VOCDataset 9 | from .wider_face import WIDERFaceDataset 10 | from .xml_style import XMLDataset 11 | 12 | from .hrsc2016 import HRSC2016Dataset 13 | from .dota import DotaDataset 14 | 15 | __all__ = [ 16 | 'CustomDataset', 'XMLDataset', 'CocoDataset', 'VOCDataset', 17 | 'CityscapesDataset', 'GroupSampler', 'DistributedGroupSampler', 18 | 'build_dataloader', 'ConcatDataset', 'RepeatDataset', 19 | 'WIDERFaceDataset', 'DATASETS', 'build_dataset', 20 | 'HRSC2016Dataset','DotaDataset' 21 | ] 22 | -------------------------------------------------------------------------------- /mmdet/datasets/builder.py: -------------------------------------------------------------------------------- 1 | import copy 2 | 3 | from mmdet.utils import build_from_cfg 4 | from .dataset_wrappers import ConcatDataset, RepeatDataset 5 | from .registry import DATASETS 6 | 7 | 8 | def _concat_dataset(cfg, default_args=None): 9 | ann_files = cfg['ann_file'] 10 | img_prefixes = cfg.get('img_prefix', None) 11 | seg_prefixes = cfg.get('seg_prefixes', None) 12 | proposal_files = cfg.get('proposal_file', None) 13 | 14 | datasets = [] 15 | num_dset = len(ann_files) 16 | for i in range(num_dset): 17 | data_cfg = copy.deepcopy(cfg) 18 | data_cfg['ann_file'] = ann_files[i] 19 | if isinstance(img_prefixes, (list, tuple)): 20 | data_cfg['img_prefix'] = img_prefixes[i] 21 | if isinstance(seg_prefixes, (list, tuple)): 22 | data_cfg['seg_prefix'] = seg_prefixes[i] 23 | if isinstance(proposal_files, (list, tuple)): 24 | data_cfg['proposal_file'] = proposal_files[i] 25 | datasets.append(build_dataset(data_cfg, default_args)) 26 | 27 | return ConcatDataset(datasets) 28 | 29 | 30 | def build_dataset(cfg, default_args=None): 31 | if isinstance(cfg, (list, tuple)): 32 | dataset = ConcatDataset([build_dataset(c, default_args) for c in cfg]) 33 | elif cfg['type'] == 'RepeatDataset': 34 | dataset = RepeatDataset( 35 | build_dataset(cfg['dataset'], default_args), cfg['times']) 36 | elif isinstance(cfg['ann_file'], (list, tuple)): 37 | dataset = _concat_dataset(cfg, default_args) 38 | else: 39 | dataset = build_from_cfg(cfg, DATASETS, default_args) 40 | 41 | return dataset 42 | -------------------------------------------------------------------------------- /mmdet/datasets/cityscapes.py: -------------------------------------------------------------------------------- 1 | from .coco import CocoDataset 2 | from .registry import DATASETS 3 | 4 | 5 | @DATASETS.register_module 6 | class CityscapesDataset(CocoDataset): 7 | 8 | CLASSES = ('person', 'rider', 'car', 'truck', 'bus', 'train', 'motorcycle', 9 | 'bicycle') 10 | -------------------------------------------------------------------------------- /mmdet/datasets/dataset_wrappers.py: -------------------------------------------------------------------------------- 1 | import bisect 2 | import math 3 | from collections import defaultdict 4 | 5 | import numpy as np 6 | from torch.utils.data.dataset import ConcatDataset as _ConcatDataset 7 | 8 | from .registry import DATASETS 9 | 10 | 11 | @DATASETS.register_module 12 | class ConcatDataset(_ConcatDataset): 13 | """A wrapper of concatenated dataset. 14 | 15 | Same as :obj:`torch.utils.data.dataset.ConcatDataset`, but 16 | concat the group flag for image aspect ratio. 17 | 18 | Args: 19 | datasets (list[:obj:`Dataset`]): A list of datasets. 20 | """ 21 | 22 | def __init__(self, datasets): 23 | super(ConcatDataset, self).__init__(datasets) 24 | self.CLASSES = datasets[0].CLASSES 25 | if hasattr(datasets[0], 'flag'): 26 | flags = [] 27 | for i in range(0, len(datasets)): 28 | flags.append(datasets[i].flag) 29 | self.flag = np.concatenate(flags) 30 | 31 | def get_cat_ids(self, idx): 32 | if idx < 0: 33 | if -idx > len(self): 34 | raise ValueError( 35 | 'absolute value of index should not exceed dataset length') 36 | idx = len(self) + idx 37 | dataset_idx = bisect.bisect_right(self.cumulative_sizes, idx) 38 | if dataset_idx == 0: 39 | sample_idx = idx 40 | else: 41 | sample_idx = idx - self.cumulative_sizes[dataset_idx - 1] 42 | return self.datasets[dataset_idx].get_cat_ids(sample_idx) 43 | 44 | 45 | @DATASETS.register_module 46 | class RepeatDataset(object): 47 | """A wrapper of repeated dataset. 48 | 49 | The length of repeated dataset will be `times` larger than the original 50 | dataset. This is useful when the data loading time is long but the dataset 51 | is small. Using RepeatDataset can reduce the data loading time between 52 | epochs. 53 | 54 | Args: 55 | dataset (:obj:`Dataset`): The dataset to be repeated. 56 | times (int): Repeat times. 57 | """ 58 | 59 | def __init__(self, dataset, times): 60 | self.dataset = dataset 61 | self.times = times 62 | self.CLASSES = dataset.CLASSES 63 | if hasattr(self.dataset, 'flag'): 64 | self.flag = np.tile(self.dataset.flag, times) 65 | 66 | self._ori_len = len(self.dataset) 67 | 68 | def __getitem__(self, idx): 69 | return self.dataset[idx % self._ori_len] 70 | 71 | def get_cat_ids(self, idx): 72 | return self.dataset.get_cat_ids(idx % self._ori_len) 73 | 74 | def __len__(self): 75 | return self.times * self._ori_len -------------------------------------------------------------------------------- /mmdet/datasets/loader/__init__.py: -------------------------------------------------------------------------------- 1 | from .build_loader import build_dataloader 2 | from .sampler import DistributedGroupSampler, GroupSampler 3 | 4 | __all__ = ['GroupSampler', 'DistributedGroupSampler', 'build_dataloader'] 5 | -------------------------------------------------------------------------------- /mmdet/datasets/loader/build_loader.py: -------------------------------------------------------------------------------- 1 | import platform 2 | from functools import partial 3 | 4 | from mmcv.parallel import collate 5 | from mmcv.runner import get_dist_info 6 | from torch.utils.data import DataLoader 7 | 8 | from .sampler import DistributedGroupSampler, DistributedSampler, GroupSampler 9 | 10 | if platform.system() != 'Windows': 11 | # https://github.com/pytorch/pytorch/issues/973 12 | import resource 13 | rlimit = resource.getrlimit(resource.RLIMIT_NOFILE) 14 | resource.setrlimit(resource.RLIMIT_NOFILE, (4096, rlimit[1])) 15 | 16 | 17 | def build_dataloader(dataset, 18 | imgs_per_gpu, 19 | workers_per_gpu, 20 | num_gpus=1, 21 | dist=True, 22 | **kwargs): 23 | shuffle = kwargs.get('shuffle', True) 24 | if dist: 25 | rank, world_size = get_dist_info() 26 | if shuffle: 27 | sampler = DistributedGroupSampler(dataset, imgs_per_gpu, 28 | world_size, rank) 29 | else: 30 | sampler = DistributedSampler( 31 | dataset, world_size, rank, shuffle=False) 32 | batch_size = imgs_per_gpu 33 | num_workers = workers_per_gpu 34 | else: 35 | sampler = GroupSampler(dataset, imgs_per_gpu) if shuffle else None 36 | batch_size = num_gpus * imgs_per_gpu 37 | num_workers = num_gpus * workers_per_gpu 38 | 39 | data_loader = DataLoader( 40 | dataset, 41 | batch_size=batch_size, 42 | sampler=sampler, 43 | num_workers=num_workers, 44 | collate_fn=partial(collate, samples_per_gpu=imgs_per_gpu), 45 | pin_memory=False, 46 | **kwargs) 47 | 48 | return data_loader 49 | -------------------------------------------------------------------------------- /mmdet/datasets/pipelines/__init__.py: -------------------------------------------------------------------------------- 1 | from .compose import Compose 2 | from .formating import (Collect, ImageToTensor, ToDataContainer, ToTensor, 3 | Transpose, to_tensor) 4 | from .loading import LoadAnnotations, LoadImageFromFile, LoadProposals 5 | from .test_aug import MultiScaleFlipAug 6 | from .transforms import (Albu, Expand, MinIoURandomCrop, Normalize, Pad, 7 | PhotoMetricDistortion, RandomCrop, RandomFlip, Resize, 8 | SegResizeFlipPadRescale) 9 | from .transforms_rotated import (PesudoRotatedRandomFlip, 10 | PesudoRotatedResize, RotatedRandomFlip, 11 | RotatedResize, RandomRotate) 12 | 13 | __all__ = [ 14 | 'Compose', 'to_tensor', 'ToTensor', 'ImageToTensor', 'ToDataContainer', 15 | 'Transpose', 'Collect', 'LoadAnnotations', 'LoadImageFromFile', 16 | 'LoadProposals', 'MultiScaleFlipAug', 'Resize', 'RandomFlip', 'Pad', 17 | 'RandomCrop', 'Normalize', 'SegResizeFlipPadRescale', 'MinIoURandomCrop', 18 | 'Expand', 'PhotoMetricDistortion', 'Albu', 'RotatedRandomFlip', 19 | 'RotatedResize', 'RandomRotate' 20 | ] 21 | -------------------------------------------------------------------------------- /mmdet/datasets/pipelines/compose.py: -------------------------------------------------------------------------------- 1 | import collections 2 | 3 | from mmdet.utils import build_from_cfg 4 | from ..registry import PIPELINES 5 | 6 | 7 | @PIPELINES.register_module 8 | class Compose(object): 9 | 10 | def __init__(self, transforms): 11 | assert isinstance(transforms, collections.abc.Sequence) 12 | self.transforms = [] 13 | for transform in transforms: 14 | if isinstance(transform, dict): 15 | transform = build_from_cfg(transform, PIPELINES) 16 | self.transforms.append(transform) 17 | elif callable(transform): 18 | self.transforms.append(transform) 19 | else: 20 | raise TypeError('transform must be callable or a dict') 21 | 22 | def __call__(self, data): 23 | for t in self.transforms: 24 | data = t(data) 25 | if data is None: 26 | return None 27 | return data 28 | 29 | def __repr__(self): 30 | format_string = self.__class__.__name__ + '(' 31 | for t in self.transforms: 32 | format_string += '\n' 33 | format_string += ' {0}'.format(t) 34 | format_string += '\n)' 35 | return format_string 36 | -------------------------------------------------------------------------------- /mmdet/datasets/pipelines/test_aug.py: -------------------------------------------------------------------------------- 1 | import mmcv 2 | 3 | from ..registry import PIPELINES 4 | from .compose import Compose 5 | 6 | 7 | @PIPELINES.register_module 8 | class MultiScaleFlipAug(object): 9 | 10 | def __init__(self, transforms, img_scale, flip=False): 11 | self.transforms = Compose(transforms) 12 | self.img_scale = img_scale if isinstance(img_scale, 13 | list) else [img_scale] 14 | assert mmcv.is_list_of(self.img_scale, tuple) 15 | self.flip = flip 16 | 17 | def __call__(self, results): 18 | aug_data = [] 19 | flip_aug = [False, True] if self.flip else [False] 20 | for scale in self.img_scale: 21 | for flip in flip_aug: 22 | _results = results.copy() 23 | _results['scale'] = scale 24 | _results['flip'] = flip 25 | data = self.transforms(_results) 26 | aug_data.append(data) 27 | # list of dict to dict of list 28 | aug_data_dict = {key: [] for key in aug_data[0]} 29 | for data in aug_data: 30 | for key, val in data.items(): 31 | aug_data_dict[key].append(val) 32 | return aug_data_dict 33 | 34 | def __repr__(self): 35 | repr_str = self.__class__.__name__ 36 | repr_str += '(transforms={}, img_scale={}, flip={})'.format( 37 | self.transforms, self.img_scale, self.flip) 38 | return repr_str 39 | -------------------------------------------------------------------------------- /mmdet/datasets/registry.py: -------------------------------------------------------------------------------- 1 | from mmdet.utils import Registry 2 | 3 | DATASETS = Registry('dataset') 4 | PIPELINES = Registry('pipeline') 5 | -------------------------------------------------------------------------------- /mmdet/datasets/voc.py: -------------------------------------------------------------------------------- 1 | from .registry import DATASETS 2 | from .xml_style import XMLDataset 3 | 4 | 5 | @DATASETS.register_module 6 | class VOCDataset(XMLDataset): 7 | 8 | CLASSES = ('aeroplane', 'bicycle', 'bird', 'boat', 'bottle', 'bus', 'car', 9 | 'cat', 'chair', 'cow', 'diningtable', 'dog', 'horse', 10 | 'motorbike', 'person', 'pottedplant', 'sheep', 'sofa', 'train', 11 | 'tvmonitor') 12 | 13 | def __init__(self, **kwargs): 14 | super(VOCDataset, self).__init__(**kwargs) 15 | if 'VOC2007' in self.img_prefix: 16 | self.year = 2007 17 | elif 'VOC2012' in self.img_prefix: 18 | self.year = 2012 19 | else: 20 | raise ValueError('Cannot infer dataset year from img_prefix') 21 | -------------------------------------------------------------------------------- /mmdet/datasets/wider_face.py: -------------------------------------------------------------------------------- 1 | import os.path as osp 2 | import xml.etree.ElementTree as ET 3 | 4 | import mmcv 5 | 6 | from .registry import DATASETS 7 | from .xml_style import XMLDataset 8 | 9 | 10 | @DATASETS.register_module 11 | class WIDERFaceDataset(XMLDataset): 12 | """ 13 | Reader for the WIDER Face dataset in PASCAL VOC format. 14 | Conversion scripts can be found in 15 | https://github.com/sovrasov/wider-face-pascal-voc-annotations 16 | """ 17 | CLASSES = ('face', ) 18 | 19 | def __init__(self, **kwargs): 20 | super(WIDERFaceDataset, self).__init__(**kwargs) 21 | 22 | def load_annotations(self, ann_file): 23 | img_infos = [] 24 | img_ids = mmcv.list_from_file(ann_file) 25 | for img_id in img_ids: 26 | filename = '{}.jpg'.format(img_id) 27 | xml_path = osp.join(self.img_prefix, 'Annotations', 28 | '{}.xml'.format(img_id)) 29 | tree = ET.parse(xml_path) 30 | root = tree.getroot() 31 | size = root.find('size') 32 | width = int(size.find('width').text) 33 | height = int(size.find('height').text) 34 | folder = root.find('folder').text 35 | img_infos.append( 36 | dict( 37 | id=img_id, 38 | filename=osp.join(folder, filename), 39 | width=width, 40 | height=height)) 41 | 42 | return img_infos 43 | -------------------------------------------------------------------------------- /mmdet/models/__init__.py: -------------------------------------------------------------------------------- 1 | from .anchor_heads import * # noqa: F401,F403 2 | from .backbones import * # noqa: F401,F403 3 | from .bbox_heads import * # noqa: F401,F403 4 | from .builder import (build_backbone, build_detector, build_head, build_loss, 5 | build_neck, build_roi_extractor, build_shared_head) 6 | from .detectors import * # noqa: F401,F403 7 | from .losses import * # noqa: F401,F403 8 | from .mask_heads import * # noqa: F401,F403 9 | from .necks import * # noqa: F401,F403 10 | from .registry import (BACKBONES, DETECTORS, HEADS, LOSSES, NECKS, 11 | ROI_EXTRACTORS, SHARED_HEADS) 12 | from .roi_extractors import * # noqa: F401,F403 13 | from .shared_heads import * # noqa: F401,F403 14 | 15 | from .bbox_heads_rotated import * 16 | 17 | from .anchor_heads_rotated import * 18 | 19 | __all__ = [ 20 | 'BACKBONES', 'NECKS', 'ROI_EXTRACTORS', 'SHARED_HEADS', 'HEADS', 'LOSSES', 21 | 'DETECTORS', 'build_backbone', 'build_neck', 'build_roi_extractor', 22 | 'build_shared_head', 'build_head', 'build_loss', 'build_detector' 23 | ] 24 | -------------------------------------------------------------------------------- /mmdet/models/anchor_heads/__init__.py: -------------------------------------------------------------------------------- 1 | from .anchor_head import AnchorHead 2 | from .fcos_head import FCOSHead 3 | from .fovea_head import FoveaHead 4 | from .fsaf_head import FSAFHead 5 | from .ga_retina_head import GARetinaHead 6 | from .ga_rpn_head import GARPNHead 7 | from .guided_anchor_head import FeatureAdaption, GuidedAnchorHead 8 | from .reppoints_head import RepPointsHead 9 | from .retina_head import RetinaHead 10 | from .rpn_head import RPNHead 11 | from .ssd_head import SSDHead 12 | 13 | __all__ = [ 14 | 'AnchorHead', 'GuidedAnchorHead', 'FeatureAdaption', 'RPNHead', 15 | 'GARPNHead', 'RetinaHead', 'GARetinaHead', 'SSDHead', 'FCOSHead', 16 | 'RepPointsHead', 'FoveaHead', 'FSAFHead' 17 | ] 18 | -------------------------------------------------------------------------------- /mmdet/models/anchor_heads_rotated/__init__.py: -------------------------------------------------------------------------------- 1 | from .anchor_head_rotated import AnchorHeadRotated 2 | from .cascade_s2anet_head import CascadeS2ANetHead 3 | from .retina_head_rotated import RetinaHeadRotated 4 | from .s2anet_head import S2ANetHead 5 | 6 | __all__ = [ 7 | 'AnchorHeadRotated', 'RetinaHeadRotated', 'S2ANetHead', 'CascadeS2ANetHead' 8 | ] 9 | -------------------------------------------------------------------------------- /mmdet/models/backbones/__init__.py: -------------------------------------------------------------------------------- 1 | from .hrnet import HRNet 2 | from .resnet import ResNet, make_res_layer 3 | from .resnext import ResNeXt 4 | from .ssd_vgg import SSDVGG 5 | 6 | __all__ = ['ResNet', 'make_res_layer', 'ResNeXt', 'SSDVGG', 'HRNet'] 7 | -------------------------------------------------------------------------------- /mmdet/models/bbox_heads/__init__.py: -------------------------------------------------------------------------------- 1 | from .bbox_head import BBoxHead 2 | from .convfc_bbox_head import ConvFCBBoxHead, SharedFCBBoxHead 3 | from .double_bbox_head import DoubleConvFCBBoxHead 4 | 5 | __all__ = [ 6 | 'BBoxHead', 'ConvFCBBoxHead', 'SharedFCBBoxHead', 'DoubleConvFCBBoxHead' 7 | ] 8 | -------------------------------------------------------------------------------- /mmdet/models/bbox_heads_rotated/__init__.py: -------------------------------------------------------------------------------- 1 | from .bbox_head_rotated import BBoxHeadRotated 2 | from .convfc_bbox_head_rotated import ConvFCBBoxHeadRotated, SharedFCBBoxHeadRotated 3 | from .double_bbox_head_rotated import DoubleConvFCBBoxHeadRotated 4 | 5 | __all__ = [ 6 | 'BBoxHeadRotated', 'ConvFCBBoxHeadRotated', 'SharedFCBBoxHeadRotated', 'DoubleConvFCBBoxHeadRotated' 7 | ] 8 | -------------------------------------------------------------------------------- /mmdet/models/builder.py: -------------------------------------------------------------------------------- 1 | from torch import nn 2 | 3 | from mmdet.utils import build_from_cfg 4 | from .registry import (BACKBONES, DETECTORS, HEADS, LOSSES, NECKS, 5 | ROI_EXTRACTORS, SHARED_HEADS) 6 | 7 | 8 | def build(cfg, registry, default_args=None): 9 | if isinstance(cfg, list): 10 | modules = [ 11 | build_from_cfg(cfg_, registry, default_args) for cfg_ in cfg 12 | ] 13 | return nn.Sequential(*modules) 14 | else: 15 | return build_from_cfg(cfg, registry, default_args) 16 | 17 | 18 | def build_backbone(cfg): 19 | return build(cfg, BACKBONES) 20 | 21 | 22 | def build_neck(cfg): 23 | return build(cfg, NECKS) 24 | 25 | 26 | def build_roi_extractor(cfg): 27 | return build(cfg, ROI_EXTRACTORS) 28 | 29 | 30 | def build_shared_head(cfg): 31 | return build(cfg, SHARED_HEADS) 32 | 33 | 34 | def build_head(cfg): 35 | return build(cfg, HEADS) 36 | 37 | 38 | def build_loss(cfg): 39 | return build(cfg, LOSSES) 40 | 41 | 42 | def build_detector(cfg, train_cfg=None, test_cfg=None): 43 | return build(cfg, DETECTORS, dict(train_cfg=train_cfg, test_cfg=test_cfg)) 44 | -------------------------------------------------------------------------------- /mmdet/models/detectors/__init__.py: -------------------------------------------------------------------------------- 1 | from .base import BaseDetector 2 | from .cascade_rcnn import CascadeRCNN 3 | from .cascade_s2anet import CascadeS2ANetDetector 4 | from .double_head_rcnn import DoubleHeadRCNN 5 | from .fast_rcnn import FastRCNN 6 | from .faster_rcnn import FasterRCNN 7 | from .faster_rcnn_hbb_obb import FasterRCNNHBBOBB 8 | from .fcos import FCOS 9 | from .fovea import FOVEA 10 | from .grid_rcnn import GridRCNN 11 | from .htc import HybridTaskCascade 12 | from .mask_rcnn import MaskRCNN 13 | from .mask_scoring_rcnn import MaskScoringRCNN 14 | from .reppoints_detector import RepPointsDetector 15 | from .retinanet import RetinaNet 16 | from .rpn import RPN 17 | from .s2anet import S2ANetDetector 18 | from .single_stage import SingleStageDetector 19 | from .two_stage import TwoStageDetector 20 | 21 | __all__ = [ 22 | 'BaseDetector', 'SingleStageDetector', 'TwoStageDetector', 'RPN', 23 | 'FastRCNN', 'FasterRCNN', 'MaskRCNN', 'CascadeRCNN', 'HybridTaskCascade', 24 | 'DoubleHeadRCNN', 'RetinaNet', 'FCOS', 'GridRCNN', 'MaskScoringRCNN', 25 | 'RepPointsDetector', 'FOVEA', 26 | 'S2ANetDetector', 'FasterRCNNHBBOBB', 'CascadeS2ANetDetector' 27 | ] 28 | -------------------------------------------------------------------------------- /mmdet/models/detectors/fast_rcnn.py: -------------------------------------------------------------------------------- 1 | from ..registry import DETECTORS 2 | from .two_stage import TwoStageDetector 3 | 4 | 5 | @DETECTORS.register_module 6 | class FastRCNN(TwoStageDetector): 7 | 8 | def __init__(self, 9 | backbone, 10 | bbox_roi_extractor, 11 | bbox_head, 12 | train_cfg, 13 | test_cfg, 14 | neck=None, 15 | shared_head=None, 16 | mask_roi_extractor=None, 17 | mask_head=None, 18 | pretrained=None): 19 | super(FastRCNN, self).__init__( 20 | backbone=backbone, 21 | neck=neck, 22 | shared_head=shared_head, 23 | bbox_roi_extractor=bbox_roi_extractor, 24 | bbox_head=bbox_head, 25 | train_cfg=train_cfg, 26 | test_cfg=test_cfg, 27 | mask_roi_extractor=mask_roi_extractor, 28 | mask_head=mask_head, 29 | pretrained=pretrained) 30 | 31 | def forward_test(self, imgs, img_metas, proposals, **kwargs): 32 | for var, name in [(imgs, 'imgs'), (img_metas, 'img_metas')]: 33 | if not isinstance(var, list): 34 | raise TypeError('{} must be a list, but got {}'.format( 35 | name, type(var))) 36 | 37 | num_augs = len(imgs) 38 | if num_augs != len(img_metas): 39 | raise ValueError( 40 | 'num of augmentations ({}) != num of image meta ({})'.format( 41 | len(imgs), len(img_metas))) 42 | # TODO: remove the restriction of imgs_per_gpu == 1 when prepared 43 | imgs_per_gpu = imgs[0].size(0) 44 | assert imgs_per_gpu == 1 45 | 46 | if num_augs == 1: 47 | return self.simple_test(imgs[0], img_metas[0], proposals[0], 48 | **kwargs) 49 | else: 50 | return self.aug_test(imgs, img_metas, proposals, **kwargs) 51 | -------------------------------------------------------------------------------- /mmdet/models/detectors/faster_rcnn.py: -------------------------------------------------------------------------------- 1 | from ..registry import DETECTORS 2 | from .two_stage import TwoStageDetector 3 | 4 | 5 | @DETECTORS.register_module 6 | class FasterRCNN(TwoStageDetector): 7 | 8 | def __init__(self, 9 | backbone, 10 | rpn_head, 11 | bbox_roi_extractor, 12 | bbox_head, 13 | train_cfg, 14 | test_cfg, 15 | neck=None, 16 | shared_head=None, 17 | pretrained=None): 18 | super(FasterRCNN, self).__init__( 19 | backbone=backbone, 20 | neck=neck, 21 | shared_head=shared_head, 22 | rpn_head=rpn_head, 23 | bbox_roi_extractor=bbox_roi_extractor, 24 | bbox_head=bbox_head, 25 | train_cfg=train_cfg, 26 | test_cfg=test_cfg, 27 | pretrained=pretrained) 28 | -------------------------------------------------------------------------------- /mmdet/models/detectors/fcos.py: -------------------------------------------------------------------------------- 1 | from ..registry import DETECTORS 2 | from .single_stage import SingleStageDetector 3 | 4 | 5 | @DETECTORS.register_module 6 | class FCOS(SingleStageDetector): 7 | 8 | def __init__(self, 9 | backbone, 10 | neck, 11 | bbox_head, 12 | train_cfg=None, 13 | test_cfg=None, 14 | pretrained=None): 15 | super(FCOS, self).__init__(backbone, neck, bbox_head, train_cfg, 16 | test_cfg, pretrained) 17 | -------------------------------------------------------------------------------- /mmdet/models/detectors/fovea.py: -------------------------------------------------------------------------------- 1 | from ..registry import DETECTORS 2 | from .single_stage import SingleStageDetector 3 | 4 | 5 | @DETECTORS.register_module 6 | class FOVEA(SingleStageDetector): 7 | 8 | def __init__(self, 9 | backbone, 10 | neck, 11 | bbox_head, 12 | train_cfg=None, 13 | test_cfg=None, 14 | pretrained=None): 15 | super(FOVEA, self).__init__(backbone, neck, bbox_head, train_cfg, 16 | test_cfg, pretrained) 17 | -------------------------------------------------------------------------------- /mmdet/models/detectors/mask_rcnn.py: -------------------------------------------------------------------------------- 1 | from ..registry import DETECTORS 2 | from .two_stage import TwoStageDetector 3 | 4 | 5 | @DETECTORS.register_module 6 | class MaskRCNN(TwoStageDetector): 7 | 8 | def __init__(self, 9 | backbone, 10 | rpn_head, 11 | bbox_roi_extractor, 12 | bbox_head, 13 | mask_roi_extractor, 14 | mask_head, 15 | train_cfg, 16 | test_cfg, 17 | neck=None, 18 | shared_head=None, 19 | pretrained=None): 20 | super(MaskRCNN, self).__init__( 21 | backbone=backbone, 22 | neck=neck, 23 | shared_head=shared_head, 24 | rpn_head=rpn_head, 25 | bbox_roi_extractor=bbox_roi_extractor, 26 | bbox_head=bbox_head, 27 | mask_roi_extractor=mask_roi_extractor, 28 | mask_head=mask_head, 29 | train_cfg=train_cfg, 30 | test_cfg=test_cfg, 31 | pretrained=pretrained) 32 | -------------------------------------------------------------------------------- /mmdet/models/detectors/retinanet.py: -------------------------------------------------------------------------------- 1 | from ..registry import DETECTORS 2 | from .single_stage import SingleStageDetector 3 | 4 | 5 | @DETECTORS.register_module 6 | class RetinaNet(SingleStageDetector): 7 | 8 | def __init__(self, 9 | backbone, 10 | neck, 11 | bbox_head, 12 | train_cfg=None, 13 | test_cfg=None, 14 | pretrained=None): 15 | super(RetinaNet, self).__init__(backbone, neck, bbox_head, train_cfg, 16 | test_cfg, pretrained) 17 | -------------------------------------------------------------------------------- /mmdet/models/detectors/s2anet.py: -------------------------------------------------------------------------------- 1 | from .single_stage import SingleStageDetector 2 | from ..registry import DETECTORS 3 | 4 | 5 | @DETECTORS.register_module 6 | class S2ANetDetector(SingleStageDetector): 7 | 8 | def __init__(self, 9 | backbone, 10 | neck, 11 | bbox_head, 12 | train_cfg=None, 13 | test_cfg=None, 14 | pretrained=None): 15 | super(S2ANetDetector, self).__init__(backbone, neck, bbox_head, train_cfg, 16 | test_cfg, pretrained) 17 | -------------------------------------------------------------------------------- /mmdet/models/detectors/single_stage.py: -------------------------------------------------------------------------------- 1 | import torch.nn as nn 2 | 3 | from mmdet.core import bbox2result 4 | from .. import builder 5 | from ..registry import DETECTORS 6 | from .base import BaseDetector 7 | 8 | 9 | @DETECTORS.register_module 10 | class SingleStageDetector(BaseDetector): 11 | """Base class for single-stage detectors. 12 | 13 | Single-stage detectors directly and densely predict bounding boxes on the 14 | output features of the backbone+neck. 15 | """ 16 | 17 | def __init__(self, 18 | backbone, 19 | neck=None, 20 | bbox_head=None, 21 | train_cfg=None, 22 | test_cfg=None, 23 | pretrained=None): 24 | super(SingleStageDetector, self).__init__() 25 | self.backbone = builder.build_backbone(backbone) 26 | if neck is not None: 27 | self.neck = builder.build_neck(neck) 28 | self.bbox_head = builder.build_head(bbox_head) 29 | self.train_cfg = train_cfg 30 | self.test_cfg = test_cfg 31 | self.init_weights(pretrained=pretrained) 32 | 33 | def init_weights(self, pretrained=None): 34 | super(SingleStageDetector, self).init_weights(pretrained) 35 | self.backbone.init_weights(pretrained=pretrained) 36 | if self.with_neck: 37 | if isinstance(self.neck, nn.Sequential): 38 | for m in self.neck: 39 | m.init_weights() 40 | else: 41 | self.neck.init_weights() 42 | self.bbox_head.init_weights() 43 | 44 | def extract_feat(self, img): 45 | """Directly extract features from the backbone+neck 46 | """ 47 | x = self.backbone(img) 48 | if self.with_neck: 49 | x = self.neck(x) 50 | return x 51 | 52 | def forward_dummy(self, img): 53 | """Used for computing network flops. 54 | 55 | See `mmedetection/tools/get_flops.py` 56 | """ 57 | x = self.extract_feat(img) 58 | outs = self.bbox_head(x) 59 | return outs 60 | 61 | def forward_train(self, 62 | img, 63 | img_metas, 64 | gt_bboxes, 65 | gt_labels, 66 | gt_bboxes_ignore=None): 67 | x = self.extract_feat(img) 68 | outs = self.bbox_head(x) 69 | loss_inputs = outs + (gt_bboxes, gt_labels, img_metas, self.train_cfg) 70 | losses = self.bbox_head.loss( 71 | *loss_inputs, gt_bboxes_ignore=gt_bboxes_ignore) 72 | return losses 73 | 74 | def simple_test(self, img, img_meta, rescale=False): 75 | x = self.extract_feat(img) 76 | outs = self.bbox_head(x) 77 | bbox_inputs = outs + (img_meta, self.test_cfg, rescale) 78 | bbox_list = self.bbox_head.get_bboxes(*bbox_inputs) 79 | bbox_results = [ 80 | bbox2result(det_bboxes, det_labels, self.bbox_head.num_classes) 81 | for det_bboxes, det_labels in bbox_list 82 | ] 83 | return bbox_results[0] 84 | 85 | def aug_test(self, imgs, img_metas, rescale=False): 86 | raise NotImplementedError 87 | -------------------------------------------------------------------------------- /mmdet/models/losses/__init__.py: -------------------------------------------------------------------------------- 1 | from .accuracy import Accuracy, accuracy 2 | from .balanced_l1_loss import BalancedL1Loss, balanced_l1_loss 3 | from .cross_entropy_loss import (CrossEntropyLoss, binary_cross_entropy, 4 | cross_entropy, mask_cross_entropy) 5 | from .focal_loss import FocalLoss, sigmoid_focal_loss 6 | from .ghm_loss import GHMC, GHMR 7 | from .iou_loss import (BoundedIoULoss, GIoULoss, IoULoss, bounded_iou_loss, 8 | iou_loss, giou_loss) 9 | from .mse_loss import MSELoss, mse_loss 10 | from .smooth_l1_loss import SmoothL1Loss, smooth_l1_loss 11 | from .utils import reduce_loss, weight_reduce_loss, weighted_loss 12 | 13 | from .rotated_iou_loss import RotatedIoULoss 14 | 15 | __all__ = [ 16 | 'accuracy', 'Accuracy', 'cross_entropy', 'binary_cross_entropy', 17 | 'mask_cross_entropy', 'CrossEntropyLoss', 'sigmoid_focal_loss', 18 | 'FocalLoss', 'smooth_l1_loss', 'SmoothL1Loss', 'balanced_l1_loss', 19 | 'BalancedL1Loss', 'mse_loss', 'MSELoss', 'iou_loss', 'bounded_iou_loss', 20 | 'IoULoss', 'BoundedIoULoss', 'GHMC', 'GHMR', 'reduce_loss', 21 | 'weight_reduce_loss', 'weighted_loss', 'GIoULoss', 'RotatedIoULoss' 22 | ] 23 | -------------------------------------------------------------------------------- /mmdet/models/losses/accuracy.py: -------------------------------------------------------------------------------- 1 | import torch.nn as nn 2 | 3 | 4 | def accuracy(pred, target, topk=1): 5 | assert isinstance(topk, (int, tuple)) 6 | if isinstance(topk, int): 7 | topk = (topk, ) 8 | return_single = True 9 | else: 10 | return_single = False 11 | 12 | maxk = max(topk) 13 | _, pred_label = pred.topk(maxk, dim=1) 14 | pred_label = pred_label.t() 15 | correct = pred_label.eq(target.view(1, -1).expand_as(pred_label)) 16 | 17 | res = [] 18 | for k in topk: 19 | correct_k = correct[:k].view(-1).float().sum(0, keepdim=True) 20 | res.append(correct_k.mul_(100.0 / pred.size(0))) 21 | return res[0] if return_single else res 22 | 23 | 24 | class Accuracy(nn.Module): 25 | 26 | def __init__(self, topk=(1, )): 27 | super().__init__() 28 | self.topk = topk 29 | 30 | def forward(self, pred, target): 31 | return accuracy(pred, target, self.topk) 32 | -------------------------------------------------------------------------------- /mmdet/models/losses/balanced_l1_loss.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import torch 3 | import torch.nn as nn 4 | 5 | from ..registry import LOSSES 6 | from .utils import weighted_loss 7 | 8 | 9 | @weighted_loss 10 | def balanced_l1_loss(pred, 11 | target, 12 | beta=1.0, 13 | alpha=0.5, 14 | gamma=1.5, 15 | reduction='mean'): 16 | assert beta > 0 17 | assert pred.size() == target.size() and target.numel() > 0 18 | 19 | diff = torch.abs(pred - target) 20 | b = np.e**(gamma / alpha) - 1 21 | loss = torch.where( 22 | diff < beta, alpha / b * 23 | (b * diff + 1) * torch.log(b * diff / beta + 1) - alpha * diff, 24 | gamma * diff + gamma / b - alpha * beta) 25 | 26 | return loss 27 | 28 | 29 | @LOSSES.register_module 30 | class BalancedL1Loss(nn.Module): 31 | """Balanced L1 Loss 32 | 33 | arXiv: https://arxiv.org/pdf/1904.02701.pdf (CVPR 2019) 34 | """ 35 | 36 | def __init__(self, 37 | alpha=0.5, 38 | gamma=1.5, 39 | beta=1.0, 40 | reduction='mean', 41 | loss_weight=1.0): 42 | super(BalancedL1Loss, self).__init__() 43 | self.alpha = alpha 44 | self.gamma = gamma 45 | self.beta = beta 46 | self.reduction = reduction 47 | self.loss_weight = loss_weight 48 | 49 | def forward(self, 50 | pred, 51 | target, 52 | weight=None, 53 | avg_factor=None, 54 | reduction_override=None, 55 | **kwargs): 56 | assert reduction_override in (None, 'none', 'mean', 'sum') 57 | reduction = ( 58 | reduction_override if reduction_override else self.reduction) 59 | loss_bbox = self.loss_weight * balanced_l1_loss( 60 | pred, 61 | target, 62 | weight, 63 | alpha=self.alpha, 64 | gamma=self.gamma, 65 | beta=self.beta, 66 | reduction=reduction, 67 | avg_factor=avg_factor, 68 | **kwargs) 69 | return loss_bbox 70 | -------------------------------------------------------------------------------- /mmdet/models/losses/focal_loss.py: -------------------------------------------------------------------------------- 1 | import torch.nn as nn 2 | import torch.nn.functional as F 3 | 4 | from mmdet.ops import sigmoid_focal_loss as _sigmoid_focal_loss 5 | from ..registry import LOSSES 6 | from .utils import weight_reduce_loss 7 | 8 | 9 | # This method is only for debugging 10 | def py_sigmoid_focal_loss(pred, 11 | target, 12 | weight=None, 13 | gamma=2.0, 14 | alpha=0.25, 15 | reduction='mean', 16 | avg_factor=None): 17 | pred_sigmoid = pred.sigmoid() 18 | target = target.type_as(pred) 19 | pt = (1 - pred_sigmoid) * target + pred_sigmoid * (1 - target) 20 | focal_weight = (alpha * target + (1 - alpha) * 21 | (1 - target)) * pt.pow(gamma) 22 | loss = F.binary_cross_entropy_with_logits( 23 | pred, target, reduction='none') * focal_weight 24 | loss = weight_reduce_loss(loss, weight, reduction, avg_factor) 25 | return loss 26 | 27 | 28 | def sigmoid_focal_loss(pred, 29 | target, 30 | weight=None, 31 | gamma=2.0, 32 | alpha=0.25, 33 | reduction='mean', 34 | avg_factor=None): 35 | # Function.apply does not accept keyword arguments, so the decorator 36 | # "weighted_loss" is not applicable 37 | loss = _sigmoid_focal_loss(pred, target, gamma, alpha) 38 | # TODO: find a proper way to handle the shape of weight 39 | if weight is not None: 40 | weight = weight.view(-1, 1) 41 | loss = weight_reduce_loss(loss, weight, reduction, avg_factor) 42 | return loss 43 | 44 | @LOSSES.register_module 45 | class FocalLoss(nn.Module): 46 | 47 | def __init__(self, 48 | use_sigmoid=True, 49 | gamma=2.0, 50 | alpha=0.25, 51 | reduction='mean', 52 | loss_weight=1.0): 53 | super(FocalLoss, self).__init__() 54 | assert use_sigmoid is True, 'Only sigmoid focal loss supported now.' 55 | self.use_sigmoid = use_sigmoid 56 | self.gamma = gamma 57 | self.alpha = alpha 58 | self.reduction = reduction 59 | self.loss_weight = loss_weight 60 | 61 | def forward(self, 62 | pred, 63 | target, 64 | weight=None, 65 | avg_factor=None, 66 | reduction_override=None): 67 | assert reduction_override in (None, 'none', 'mean', 'sum') 68 | reduction = ( 69 | reduction_override if reduction_override else self.reduction) 70 | if self.use_sigmoid: 71 | loss_cls = self.loss_weight * sigmoid_focal_loss( 72 | pred, 73 | target, 74 | weight, 75 | gamma=self.gamma, 76 | alpha=self.alpha, 77 | reduction=reduction, 78 | avg_factor=avg_factor) 79 | else: 80 | raise NotImplementedError 81 | return loss_cls 82 | -------------------------------------------------------------------------------- /mmdet/models/losses/mse_loss.py: -------------------------------------------------------------------------------- 1 | import torch.nn as nn 2 | import torch.nn.functional as F 3 | 4 | from ..registry import LOSSES 5 | from .utils import weighted_loss 6 | 7 | mse_loss = weighted_loss(F.mse_loss) 8 | 9 | 10 | @LOSSES.register_module 11 | class MSELoss(nn.Module): 12 | 13 | def __init__(self, reduction='mean', loss_weight=1.0): 14 | super().__init__() 15 | self.reduction = reduction 16 | self.loss_weight = loss_weight 17 | 18 | def forward(self, pred, target, weight=None, avg_factor=None): 19 | loss = self.loss_weight * mse_loss( 20 | pred, 21 | target, 22 | weight, 23 | reduction=self.reduction, 24 | avg_factor=avg_factor) 25 | return loss 26 | -------------------------------------------------------------------------------- /mmdet/models/losses/rotated_iou_loss.py: -------------------------------------------------------------------------------- 1 | import torch 2 | import torch.nn as nn 3 | from mmdet.ops import box_iou_rotated_differentiable 4 | 5 | from ..registry import LOSSES 6 | from .utils import weighted_loss 7 | 8 | 9 | @weighted_loss 10 | def iou_loss(pred, target, linear=False, eps=1e-6): 11 | """IoU loss. 12 | 13 | Computing the IoU loss between a set of predicted bboxes and target bboxes. 14 | The loss is calculated as negative log of IoU. 15 | 16 | Args: 17 | pred (Tensor): Predicted bboxes of format (x, y, w, h, a), 18 | shape (n, 5). 19 | target (Tensor): Corresponding gt bboxes, shape (n, 5). 20 | linear (bool): If True, use linear scale of loss instead of 21 | log scale. Default: False. 22 | eps (float): Eps to avoid log(0). 23 | 24 | Return: 25 | Tensor: Loss tensor. 26 | """ 27 | ious = box_iou_rotated_differentiable(pred, target).clamp(min=eps) 28 | if linear: 29 | loss = 1 - ious 30 | else: 31 | loss = -ious.log() 32 | return loss 33 | 34 | 35 | @LOSSES.register_module 36 | class RotatedIoULoss(nn.Module): 37 | 38 | def __init__(self, linear=False, eps=1e-6, reduction='mean', loss_weight=1.0): 39 | super(RotatedIoULoss, self).__init__() 40 | self.linear = linear 41 | self.eps = eps 42 | self.reduction = reduction 43 | self.loss_weight = loss_weight 44 | 45 | def forward(self, 46 | pred, 47 | target, 48 | weight=None, 49 | avg_factor=None, 50 | reduction_override=None, 51 | **kwargs): 52 | if weight is not None and not torch.any(weight > 0): 53 | return (pred * weight).sum() # 0 54 | assert reduction_override in (None, 'none', 'mean', 'sum') 55 | reduction = ( 56 | reduction_override if reduction_override else self.reduction) 57 | if (weight is not None) and (not torch.any(weight > 0)) and ( 58 | reduction != 'none'): 59 | return (pred * weight).sum() # 0 60 | if weight is not None and weight.dim() > 1: 61 | # TODO: remove this in the future 62 | # reduce the weight of shape (n, 4) to (n,) to match the 63 | # iou_loss of shape (n,) 64 | assert weight.shape == pred.shape 65 | weight = weight.mean(-1) 66 | loss = self.loss_weight * iou_loss( 67 | pred, 68 | target, 69 | weight, 70 | linear=self.linear, 71 | eps=self.eps, 72 | reduction=reduction, 73 | avg_factor=avg_factor, 74 | **kwargs) 75 | return loss 76 | -------------------------------------------------------------------------------- /mmdet/models/losses/smooth_l1_loss.py: -------------------------------------------------------------------------------- 1 | import torch 2 | import torch.nn as nn 3 | 4 | from ..registry import LOSSES 5 | from .utils import weighted_loss 6 | 7 | 8 | @weighted_loss 9 | def smooth_l1_loss(pred, target, beta=1.0): 10 | assert beta > 0 11 | assert pred.size() == target.size() and target.numel() > 0 12 | diff = torch.abs(pred - target) 13 | loss = torch.where(diff < beta, 0.5 * diff * diff / beta, 14 | diff - 0.5 * beta) 15 | return loss 16 | 17 | 18 | @LOSSES.register_module 19 | class SmoothL1Loss(nn.Module): 20 | 21 | def __init__(self, beta=1.0, reduction='mean', loss_weight=1.0): 22 | super(SmoothL1Loss, self).__init__() 23 | self.beta = beta 24 | self.reduction = reduction 25 | self.loss_weight = loss_weight 26 | 27 | def forward(self, 28 | pred, 29 | target, 30 | weight=None, 31 | avg_factor=None, 32 | reduction_override=None, 33 | **kwargs): 34 | assert reduction_override in (None, 'none', 'mean', 'sum') 35 | reduction = ( 36 | reduction_override if reduction_override else self.reduction) 37 | loss_bbox = self.loss_weight * smooth_l1_loss( 38 | pred, 39 | target, 40 | weight, 41 | beta=self.beta, 42 | reduction=reduction, 43 | avg_factor=avg_factor, 44 | **kwargs) 45 | return loss_bbox 46 | -------------------------------------------------------------------------------- /mmdet/models/mask_heads/__init__.py: -------------------------------------------------------------------------------- 1 | from .fcn_mask_head import FCNMaskHead 2 | from .fused_semantic_head import FusedSemanticHead 3 | from .grid_head import GridHead 4 | from .htc_mask_head import HTCMaskHead 5 | from .maskiou_head import MaskIoUHead 6 | 7 | __all__ = [ 8 | 'FCNMaskHead', 'HTCMaskHead', 'FusedSemanticHead', 'GridHead', 9 | 'MaskIoUHead' 10 | ] 11 | -------------------------------------------------------------------------------- /mmdet/models/mask_heads/htc_mask_head.py: -------------------------------------------------------------------------------- 1 | from ..registry import HEADS 2 | from ..utils import ConvModule 3 | from .fcn_mask_head import FCNMaskHead 4 | 5 | 6 | @HEADS.register_module 7 | class HTCMaskHead(FCNMaskHead): 8 | 9 | def __init__(self, *args, **kwargs): 10 | super(HTCMaskHead, self).__init__(*args, **kwargs) 11 | self.conv_res = ConvModule( 12 | self.conv_out_channels, 13 | self.conv_out_channels, 14 | 1, 15 | conv_cfg=self.conv_cfg, 16 | norm_cfg=self.norm_cfg) 17 | 18 | def init_weights(self): 19 | super(HTCMaskHead, self).init_weights() 20 | self.conv_res.init_weights() 21 | 22 | def forward(self, x, res_feat=None, return_logits=True, return_feat=True): 23 | if res_feat is not None: 24 | res_feat = self.conv_res(res_feat) 25 | x = x + res_feat 26 | for conv in self.convs: 27 | x = conv(x) 28 | res_feat = x 29 | outs = [] 30 | if return_logits: 31 | x = self.upsample(x) 32 | if self.upsample_method == 'deconv': 33 | x = self.relu(x) 34 | mask_pred = self.conv_logits(x) 35 | outs.append(mask_pred) 36 | if return_feat: 37 | outs.append(res_feat) 38 | return outs if len(outs) > 1 else outs[0] 39 | -------------------------------------------------------------------------------- /mmdet/models/necks/__init__.py: -------------------------------------------------------------------------------- 1 | from .bfp import BFP 2 | from .fpn import FPN 3 | from .hrfpn import HRFPN 4 | 5 | __all__ = ['FPN', 'BFP', 'HRFPN'] 6 | -------------------------------------------------------------------------------- /mmdet/models/plugins/__init__.py: -------------------------------------------------------------------------------- 1 | from .generalized_attention import GeneralizedAttention 2 | from .non_local import NonLocal2D 3 | 4 | __all__ = ['NonLocal2D', 'GeneralizedAttention'] 5 | -------------------------------------------------------------------------------- /mmdet/models/registry.py: -------------------------------------------------------------------------------- 1 | from mmdet.utils import Registry 2 | 3 | BACKBONES = Registry('backbone') 4 | NECKS = Registry('neck') 5 | ROI_EXTRACTORS = Registry('roi_extractor') 6 | SHARED_HEADS = Registry('shared_head') 7 | HEADS = Registry('head') 8 | LOSSES = Registry('loss') 9 | DETECTORS = Registry('detector') 10 | -------------------------------------------------------------------------------- /mmdet/models/roi_extractors/__init__.py: -------------------------------------------------------------------------------- 1 | from .single_level import SingleRoIExtractor 2 | from .single_level_rotated import SingleRoIExtractorRotated 3 | 4 | __all__ = ['SingleRoIExtractor', 'SingleRoIExtractorRotated'] 5 | -------------------------------------------------------------------------------- /mmdet/models/roi_extractors/single_level_rotated.py: -------------------------------------------------------------------------------- 1 | from __future__ import division 2 | 3 | import torch 4 | 5 | from .single_level import SingleRoIExtractor 6 | from ..registry import ROI_EXTRACTORS 7 | 8 | 9 | @ROI_EXTRACTORS.register_module 10 | class SingleRoIExtractorRotated(SingleRoIExtractor): 11 | 12 | def map_roi_levels(self, rois, num_levels): 13 | """Map rois to corresponding feature levels by scales. 14 | 15 | - scale < finest_scale * 2: level 0 16 | - finest_scale * 2 <= scale < finest_scale * 4: level 1 17 | - finest_scale * 4 <= scale < finest_scale * 8: level 2 18 | - scale >= finest_scale * 8: level 3 19 | 20 | Args: 21 | rois (Tensor): Input RoIs, shape (k, 5). 22 | num_levels (int): Total level number. 23 | 24 | Returns: 25 | Tensor: Level index (0-based) of each RoI, shape (k, ) 26 | """ 27 | scale = torch.sqrt((rois[:, 3] + 1) * (rois[:, 4] + 1)) 28 | target_lvls = torch.floor(torch.log2(scale / self.finest_scale + 1e-6)) 29 | target_lvls = target_lvls.clamp(min=0, max=num_levels - 1).long() 30 | return target_lvls 31 | 32 | def roi_rescale(self, rois, scale_factor): 33 | cx = rois[:, 1] 34 | cy = rois[:, 2] 35 | w = rois[:, 3] + 1 36 | h = rois[:, 4] + 1 37 | a = rois[:, 5] 38 | new_w = w * scale_factor 39 | new_h = h * scale_factor 40 | new_rois = torch.stack((rois[:, 0], cx, cy, new_w, new_h, a), dim=-1) 41 | return new_rois 42 | -------------------------------------------------------------------------------- /mmdet/models/shared_heads/__init__.py: -------------------------------------------------------------------------------- 1 | from .res_layer import ResLayer 2 | 3 | __all__ = ['ResLayer'] 4 | -------------------------------------------------------------------------------- /mmdet/models/shared_heads/res_layer.py: -------------------------------------------------------------------------------- 1 | import logging 2 | 3 | import torch.nn as nn 4 | from mmcv.cnn import constant_init, kaiming_init 5 | from mmcv.runner import load_checkpoint 6 | 7 | from mmdet.core import auto_fp16 8 | from ..backbones import ResNet, make_res_layer 9 | from ..registry import SHARED_HEADS 10 | 11 | 12 | @SHARED_HEADS.register_module 13 | class ResLayer(nn.Module): 14 | 15 | def __init__(self, 16 | depth, 17 | stage=3, 18 | stride=2, 19 | dilation=1, 20 | style='pytorch', 21 | norm_cfg=dict(type='BN', requires_grad=True), 22 | norm_eval=True, 23 | with_cp=False, 24 | dcn=None): 25 | super(ResLayer, self).__init__() 26 | self.norm_eval = norm_eval 27 | self.norm_cfg = norm_cfg 28 | self.stage = stage 29 | self.fp16_enabled = False 30 | block, stage_blocks = ResNet.arch_settings[depth] 31 | stage_block = stage_blocks[stage] 32 | planes = 64 * 2**stage 33 | inplanes = 64 * 2**(stage - 1) * block.expansion 34 | 35 | res_layer = make_res_layer( 36 | block, 37 | inplanes, 38 | planes, 39 | stage_block, 40 | stride=stride, 41 | dilation=dilation, 42 | style=style, 43 | with_cp=with_cp, 44 | norm_cfg=self.norm_cfg, 45 | dcn=dcn) 46 | self.add_module('layer{}'.format(stage + 1), res_layer) 47 | 48 | def init_weights(self, pretrained=None): 49 | if isinstance(pretrained, str): 50 | logger = logging.getLogger() 51 | load_checkpoint(self, pretrained, strict=False, logger=logger) 52 | elif pretrained is None: 53 | for m in self.modules(): 54 | if isinstance(m, nn.Conv2d): 55 | kaiming_init(m) 56 | elif isinstance(m, nn.BatchNorm2d): 57 | constant_init(m, 1) 58 | else: 59 | raise TypeError('pretrained must be a str or None') 60 | 61 | @auto_fp16() 62 | def forward(self, x): 63 | res_layer = getattr(self, 'layer{}'.format(self.stage + 1)) 64 | out = res_layer(x) 65 | return out 66 | 67 | def train(self, mode=True): 68 | super(ResLayer, self).train(mode) 69 | if self.norm_eval: 70 | for m in self.modules(): 71 | if isinstance(m, nn.BatchNorm2d): 72 | m.eval() 73 | -------------------------------------------------------------------------------- /mmdet/models/utils/__init__.py: -------------------------------------------------------------------------------- 1 | from .conv_module import ConvModule, build_conv_layer 2 | from .conv_ws import ConvWS2d, conv_ws_2d 3 | from .norm import build_norm_layer 4 | from .scale import Scale 5 | from .weight_init import (bias_init_with_prob, kaiming_init, normal_init, 6 | uniform_init, xavier_init) 7 | 8 | __all__ = [ 9 | 'conv_ws_2d', 'ConvWS2d', 'build_conv_layer', 'ConvModule', 10 | 'build_norm_layer', 'xavier_init', 'normal_init', 'uniform_init', 11 | 'kaiming_init', 'bias_init_with_prob', 'Scale' 12 | ] 13 | -------------------------------------------------------------------------------- /mmdet/models/utils/conv_ws.py: -------------------------------------------------------------------------------- 1 | import torch.nn as nn 2 | import torch.nn.functional as F 3 | 4 | 5 | def conv_ws_2d(input, 6 | weight, 7 | bias=None, 8 | stride=1, 9 | padding=0, 10 | dilation=1, 11 | groups=1, 12 | eps=1e-5): 13 | c_in = weight.size(0) 14 | weight_flat = weight.view(c_in, -1) 15 | mean = weight_flat.mean(dim=1, keepdim=True).view(c_in, 1, 1, 1) 16 | std = weight_flat.std(dim=1, keepdim=True).view(c_in, 1, 1, 1) 17 | weight = (weight - mean) / (std + eps) 18 | return F.conv2d(input, weight, bias, stride, padding, dilation, groups) 19 | 20 | 21 | class ConvWS2d(nn.Conv2d): 22 | 23 | def __init__(self, 24 | in_channels, 25 | out_channels, 26 | kernel_size, 27 | stride=1, 28 | padding=0, 29 | dilation=1, 30 | groups=1, 31 | bias=True, 32 | eps=1e-5): 33 | super(ConvWS2d, self).__init__( 34 | in_channels, 35 | out_channels, 36 | kernel_size, 37 | stride=stride, 38 | padding=padding, 39 | dilation=dilation, 40 | groups=groups, 41 | bias=bias) 42 | self.eps = eps 43 | 44 | def forward(self, x): 45 | return conv_ws_2d(x, self.weight, self.bias, self.stride, self.padding, 46 | self.dilation, self.groups, self.eps) 47 | -------------------------------------------------------------------------------- /mmdet/models/utils/norm.py: -------------------------------------------------------------------------------- 1 | import torch.nn as nn 2 | 3 | norm_cfg = { 4 | # format: layer_type: (abbreviation, module) 5 | 'BN': ('bn', nn.BatchNorm2d), 6 | 'SyncBN': ('bn', nn.SyncBatchNorm), 7 | 'GN': ('gn', nn.GroupNorm), 8 | # and potentially 'SN' 9 | } 10 | 11 | 12 | def build_norm_layer(cfg, num_features, postfix=''): 13 | """ Build normalization layer 14 | 15 | Args: 16 | cfg (dict): cfg should contain: 17 | type (str): identify norm layer type. 18 | layer args: args needed to instantiate a norm layer. 19 | requires_grad (bool): [optional] whether stop gradient updates 20 | num_features (int): number of channels from input. 21 | postfix (int, str): appended into norm abbreviation to 22 | create named layer. 23 | 24 | Returns: 25 | name (str): abbreviation + postfix 26 | layer (nn.Module): created norm layer 27 | """ 28 | assert isinstance(cfg, dict) and 'type' in cfg 29 | cfg_ = cfg.copy() 30 | 31 | layer_type = cfg_.pop('type') 32 | if layer_type not in norm_cfg: 33 | raise KeyError('Unrecognized norm type {}'.format(layer_type)) 34 | else: 35 | abbr, norm_layer = norm_cfg[layer_type] 36 | if norm_layer is None: 37 | raise NotImplementedError 38 | 39 | assert isinstance(postfix, (int, str)) 40 | name = abbr + str(postfix) 41 | 42 | requires_grad = cfg_.pop('requires_grad', True) 43 | cfg_.setdefault('eps', 1e-5) 44 | if layer_type != 'GN': 45 | layer = norm_layer(num_features, **cfg_) 46 | if layer_type == 'SyncBN': 47 | layer._specify_ddp_gpu_num(1) 48 | else: 49 | assert 'num_groups' in cfg_ 50 | layer = norm_layer(num_channels=num_features, **cfg_) 51 | 52 | for param in layer.parameters(): 53 | param.requires_grad = requires_grad 54 | 55 | return name, layer 56 | -------------------------------------------------------------------------------- /mmdet/models/utils/scale.py: -------------------------------------------------------------------------------- 1 | import torch 2 | import torch.nn as nn 3 | 4 | 5 | class Scale(nn.Module): 6 | """ 7 | A learnable scale parameter 8 | """ 9 | 10 | def __init__(self, scale=1.0): 11 | super(Scale, self).__init__() 12 | self.scale = nn.Parameter(torch.tensor(scale, dtype=torch.float)) 13 | 14 | def forward(self, x): 15 | return x * self.scale 16 | -------------------------------------------------------------------------------- /mmdet/models/utils/weight_init.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import torch.nn as nn 3 | 4 | 5 | def xavier_init(module, gain=1, bias=0, distribution='normal'): 6 | assert distribution in ['uniform', 'normal'] 7 | if distribution == 'uniform': 8 | nn.init.xavier_uniform_(module.weight, gain=gain) 9 | else: 10 | nn.init.xavier_normal_(module.weight, gain=gain) 11 | if hasattr(module, 'bias'): 12 | nn.init.constant_(module.bias, bias) 13 | 14 | 15 | def normal_init(module, mean=0, std=1, bias=0): 16 | nn.init.normal_(module.weight, mean, std) 17 | if hasattr(module, 'bias'): 18 | nn.init.constant_(module.bias, bias) 19 | 20 | 21 | def uniform_init(module, a=0, b=1, bias=0): 22 | nn.init.uniform_(module.weight, a, b) 23 | if hasattr(module, 'bias'): 24 | nn.init.constant_(module.bias, bias) 25 | 26 | 27 | def kaiming_init(module, 28 | mode='fan_out', 29 | nonlinearity='relu', 30 | bias=0, 31 | distribution='normal'): 32 | assert distribution in ['uniform', 'normal'] 33 | if distribution == 'uniform': 34 | nn.init.kaiming_uniform_( 35 | module.weight, mode=mode, nonlinearity=nonlinearity) 36 | else: 37 | nn.init.kaiming_normal_( 38 | module.weight, mode=mode, nonlinearity=nonlinearity) 39 | if hasattr(module, 'bias'): 40 | nn.init.constant_(module.bias, bias) 41 | 42 | 43 | def bias_init_with_prob(prior_prob): 44 | """ initialize conv/fc bias value according to giving probablity""" 45 | bias_init = float(-np.log((1 - prior_prob) / prior_prob)) 46 | return bias_init 47 | -------------------------------------------------------------------------------- /mmdet/ops/__init__.py: -------------------------------------------------------------------------------- 1 | from .context_block import ContextBlock 2 | from .dcn import (DeformConv, DeformConvPack, DeformRoIPooling, 3 | DeformRoIPoolingPack, ModulatedDeformConv, 4 | ModulatedDeformConvPack, ModulatedDeformRoIPoolingPack, 5 | deform_conv, deform_roi_pooling, modulated_deform_conv) 6 | from .masked_conv import MaskedConv2d 7 | from .nms import nms, soft_nms 8 | from .nms_rotated import nms_rotated 9 | from .roi_align import RoIAlign, roi_align 10 | from .roi_align_rotated import RoIAlignRotated 11 | from .roi_pool import RoIPool, roi_pool 12 | from .sigmoid_focal_loss import SigmoidFocalLoss, sigmoid_focal_loss 13 | from .ml_nms_rotated import ml_nms_rotated 14 | 15 | from .box_iou_rotated_diff import box_iou_rotated_differentiable 16 | 17 | __all__ = [ 18 | 'nms', 'soft_nms', 'RoIAlign', 'roi_align', 'RoIPool', 'roi_pool', 19 | 'DeformConv', 'DeformConvPack', 'DeformRoIPooling', 'DeformRoIPoolingPack', 20 | 'ModulatedDeformRoIPoolingPack', 'ModulatedDeformConv', 21 | 'ModulatedDeformConvPack', 'deform_conv', 'modulated_deform_conv', 22 | 'deform_roi_pooling', 'SigmoidFocalLoss', 'sigmoid_focal_loss', 23 | 'MaskedConv2d', 'ContextBlock', 24 | 'RoIAlignRotated', 'ml_nms_rotated', 'nms_rotated', 'box_iou_rotated_differentiable' 25 | ] 26 | -------------------------------------------------------------------------------- /mmdet/ops/box_iou_rotated/__init__.py: -------------------------------------------------------------------------------- 1 | from .box_iou_rotated_cuda import box_iou_rotated 2 | 3 | __all__ = ['box_iou_rotated'] 4 | -------------------------------------------------------------------------------- /mmdet/ops/box_iou_rotated/src/box_iou_rotated.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved 2 | #pragma once 3 | #include 4 | #include 5 | 6 | 7 | at::Tensor box_iou_rotated_cpu( 8 | const at::Tensor& boxes1, 9 | const at::Tensor& boxes2); 10 | 11 | #ifdef WITH_CUDA 12 | at::Tensor box_iou_rotated_cuda( 13 | const at::Tensor& boxes1, 14 | const at::Tensor& boxes2); 15 | #endif 16 | 17 | // Interface for Python 18 | // inline is needed to prevent multiple function definitions when this header is 19 | // included by different cpps 20 | inline at::Tensor box_iou_rotated( 21 | const at::Tensor& boxes1, 22 | const at::Tensor& boxes2) { 23 | assert(boxes1.device().is_cuda() == boxes2.device().is_cuda()); 24 | if (boxes1.device().is_cuda()) { 25 | #ifdef WITH_CUDA 26 | return box_iou_rotated_cuda(boxes1, boxes2); 27 | #else 28 | AT_ERROR("Not compiled with GPU support"); 29 | #endif 30 | } 31 | 32 | return box_iou_rotated_cpu(boxes1, boxes2); 33 | } 34 | 35 | 36 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { 37 | m.def("box_iou_rotated", &box_iou_rotated, "IoU for rotated boxes"); 38 | } -------------------------------------------------------------------------------- /mmdet/ops/box_iou_rotated/src/box_iou_rotated_cpu.cpp: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved 2 | #include "box_iou_rotated.h" 3 | #include "box_iou_rotated_utils.h" 4 | 5 | 6 | template 7 | void box_iou_rotated_cpu_kernel( 8 | const at::Tensor& boxes1, 9 | const at::Tensor& boxes2, 10 | at::Tensor& ious) { 11 | auto widths1 = boxes1.select(1, 2).contiguous(); 12 | auto heights1 = boxes1.select(1, 3).contiguous(); 13 | auto widths2 = boxes2.select(1, 2).contiguous(); 14 | auto heights2 = boxes2.select(1, 3).contiguous(); 15 | 16 | at::Tensor areas1 = widths1 * heights1; 17 | at::Tensor areas2 = widths2 * heights2; 18 | 19 | auto num_boxes1 = boxes1.size(0); 20 | auto num_boxes2 = boxes2.size(0); 21 | 22 | for (int i = 0; i < num_boxes1; i++) { 23 | for (int j = 0; j < num_boxes2; j++) { 24 | ious[i * num_boxes2 + j] = single_box_iou_rotated( 25 | boxes1[i].data_ptr(), boxes2[j].data_ptr()); 26 | } 27 | } 28 | } 29 | 30 | at::Tensor box_iou_rotated_cpu( 31 | const at::Tensor& boxes1, 32 | const at::Tensor& boxes2) { 33 | auto num_boxes1 = boxes1.size(0); 34 | auto num_boxes2 = boxes2.size(0); 35 | at::Tensor ious = 36 | at::empty({num_boxes1 * num_boxes2}, boxes1.options().dtype(at::kFloat)); 37 | 38 | box_iou_rotated_cpu_kernel(boxes1, boxes2, ious); 39 | 40 | // reshape from 1d array to 2d array 41 | auto shape = std::vector{num_boxes1, num_boxes2}; 42 | return ious.reshape(shape); 43 | } 44 | 45 | -------------------------------------------------------------------------------- /mmdet/ops/box_iou_rotated_diff/__init__.py: -------------------------------------------------------------------------------- 1 | from .box_iou_rotated_diff import box_iou_rotated_differentiable 2 | 3 | __all__ = ['box_iou_rotated_differentiable'] -------------------------------------------------------------------------------- /mmdet/ops/box_iou_rotated_diff/box_iou_rotated_diff.py: -------------------------------------------------------------------------------- 1 | """ 2 | Differentiable IoU calculation for rotated boxes 3 | Most of the code is adapted from https://github.com/lilanxiao/Rotated_IoU 4 | """ 5 | import torch 6 | from .box_intersection_2d import oriented_box_intersection_2d 7 | 8 | 9 | def rotated_box_to_poly(rotated_boxes: torch.Tensor): 10 | """ Transform rotated boxes to polygons 11 | Args: 12 | rotated_boxes (Tensor): (x, y, w, h, a) with shape (n, 5) 13 | Return: 14 | polys (Tensor): 4 corner points (x, y) of polygons with shape (n, 4, 2) 15 | """ 16 | cs = torch.cos(rotated_boxes[:, 4]) 17 | ss = torch.sin(rotated_boxes[:, 4]) 18 | w = rotated_boxes[:, 2] - 1 19 | h = rotated_boxes[:, 3] - 1 20 | 21 | x_ctr = rotated_boxes[:, 0] 22 | y_ctr = rotated_boxes[:, 1] 23 | x1 = x_ctr + cs * (w / 2.0) - ss * (-h / 2.0) 24 | x2 = x_ctr + cs * (w / 2.0) - ss * (h / 2.0) 25 | x3 = x_ctr + cs * (-w / 2.0) - ss * (h / 2.0) 26 | x4 = x_ctr + cs * (-w / 2.0) - ss * (-h / 2.0) 27 | 28 | y1 = y_ctr + ss * (w / 2.0) + cs * (-h / 2.0) 29 | y2 = y_ctr + ss * (w / 2.0) + cs * (h / 2.0) 30 | y3 = y_ctr + ss * (-w / 2.0) + cs * (h / 2.0) 31 | y4 = y_ctr + ss * (-w / 2.0) + cs * (-h / 2.0) 32 | 33 | polys = torch.stack([x1, y1, x2, y2, x3, y3, x4, y4], dim=-1) 34 | polys = polys.reshape(-1, 4, 2) # to (n, 4, 2) 35 | 36 | return polys 37 | 38 | 39 | def box_iou_rotated_differentiable(boxes1: torch.Tensor, boxes2: torch.Tensor, iou_only: bool = True): 40 | """Calculate IoU between rotated boxes 41 | 42 | Args: 43 | box1 (torch.Tensor): (n, 5) 44 | box2 (torch.Tensor): (n, 5) 45 | iou_only: Whether to keep other vars, e.g., polys, unions. Default True to drop these vars. 46 | 47 | Returns: 48 | iou (torch.Tensor): (n, ) 49 | polys1 (torch.Tensor): (n, 4, 2) 50 | polys2 (torch.Tensor): (n, 4, 2) 51 | U (torch.Tensor): (n) area1 + area2 - inter_area 52 | """ 53 | # transform to polygons 54 | polys1 = rotated_box_to_poly(boxes1) 55 | polys2 = rotated_box_to_poly(boxes2) 56 | # calculate insection areas 57 | inter_area, _ = oriented_box_intersection_2d(polys1, polys2) 58 | area1 = boxes1[..., 2] * boxes1[..., 3] 59 | area2 = boxes2[..., 2] * boxes2[..., 3] 60 | union = area1 + area2 - inter_area 61 | iou = inter_area / union 62 | if iou_only: 63 | return iou 64 | else: 65 | return iou, union, polys1, polys2, 66 | -------------------------------------------------------------------------------- /mmdet/ops/box_iou_rotated_diff/src/cuda_utils.h: -------------------------------------------------------------------------------- 1 | #ifndef _CUDA_UTILS_H 2 | #define _CUDA_UTILS_H 3 | 4 | #include 5 | #include 6 | #include 7 | #include 8 | #include 9 | #include 10 | 11 | #define TOTAL_THREADS 512 12 | 13 | inline int opt_n_thread(int work_size){ 14 | const int pow_2 = std::log(static_cast(work_size)) / std::log(2.0); 15 | return max(min(1<(), mask.data_ptr(), 24 | num_valid.data_ptr(), idx.data_ptr()); 25 | 26 | return idx; 27 | } 28 | 29 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m){ 30 | m.def("sort_vertices_forward", &sort_vertices, "sort vertices of a convex polygon. forward only"); 31 | } -------------------------------------------------------------------------------- /mmdet/ops/box_iou_rotated_diff/src/sort_vert.h: -------------------------------------------------------------------------------- 1 | #pragma once 2 | #include 3 | 4 | #define MAX_NUM_VERT_IDX 9 5 | 6 | at::Tensor sort_vertices(at::Tensor vertices, at::Tensor mask, at::Tensor num_valid); -------------------------------------------------------------------------------- /mmdet/ops/box_iou_rotated_diff/src/utils.h: -------------------------------------------------------------------------------- 1 | #pragma once 2 | #include 3 | #include 4 | 5 | #define CHECK_CUDA(x) \ 6 | do { \ 7 | TORCH_CHECK(x.is_cuda(), #x " must be a CUDA tensor"); \ 8 | } while (0) 9 | 10 | #define CHECK_CONTIGUOUS(x) \ 11 | do { \ 12 | TORCH_CHECK(x.is_contiguous(), #x " must ne a contiguous tensor"); \ 13 | } while (0) 14 | 15 | #define CHECK_IS_INT(x) \ 16 | do { \ 17 | TORCH_CHECK(x.scalar_type()==at::ScalarType::Int, \ 18 | #x " must be a int tensor"); \ 19 | } while (0) 20 | 21 | #define CHECK_IS_FLOAT(x) \ 22 | do { \ 23 | TORCH_CHECK(x.scalar_type()==at::ScalarType::Float, \ 24 | #x " must be a float tensor"); \ 25 | } while (0) 26 | 27 | #define CHECK_IS_BOOL(x) \ 28 | do { \ 29 | TORCH_CHECK(x.scalar_type()==at::ScalarType::Bool, \ 30 | #x " must be a bool tensor"); \ 31 | } while (0) -------------------------------------------------------------------------------- /mmdet/ops/dcn/__init__.py: -------------------------------------------------------------------------------- 1 | from .deform_conv import (DeformConv, DeformConvPack, ModulatedDeformConv, 2 | ModulatedDeformConvPack, deform_conv, 3 | modulated_deform_conv) 4 | from .deform_pool import (DeformRoIPooling, DeformRoIPoolingPack, 5 | ModulatedDeformRoIPoolingPack, deform_roi_pooling) 6 | 7 | __all__ = [ 8 | 'DeformConv', 'DeformConvPack', 'ModulatedDeformConv', 9 | 'ModulatedDeformConvPack', 'DeformRoIPooling', 'DeformRoIPoolingPack', 10 | 'ModulatedDeformRoIPoolingPack', 'deform_conv', 'modulated_deform_conv', 11 | 'deform_roi_pooling' 12 | ] 13 | -------------------------------------------------------------------------------- /mmdet/ops/masked_conv/__init__.py: -------------------------------------------------------------------------------- 1 | from .masked_conv import MaskedConv2d, masked_conv2d 2 | 3 | __all__ = ['masked_conv2d', 'MaskedConv2d'] 4 | -------------------------------------------------------------------------------- /mmdet/ops/masked_conv/src/masked_conv2d_cuda.cpp: -------------------------------------------------------------------------------- 1 | #include 2 | 3 | #include 4 | #include 5 | 6 | int MaskedIm2colForwardLaucher(const at::Tensor im, const int height, 7 | const int width, const int channels, 8 | const int kernel_h, const int kernel_w, 9 | const int pad_h, const int pad_w, 10 | const at::Tensor mask_h_idx, 11 | const at::Tensor mask_w_idx, const int mask_cnt, 12 | at::Tensor col); 13 | 14 | int MaskedCol2imForwardLaucher(const at::Tensor col, const int height, 15 | const int width, const int channels, 16 | const at::Tensor mask_h_idx, 17 | const at::Tensor mask_w_idx, const int mask_cnt, 18 | at::Tensor im); 19 | 20 | #define CHECK_CUDA(x) AT_CHECK(x.type().is_cuda(), #x, " must be a CUDAtensor ") 21 | #define CHECK_CONTIGUOUS(x) \ 22 | AT_CHECK(x.is_contiguous(), #x, " must be contiguous ") 23 | #define CHECK_INPUT(x) \ 24 | CHECK_CUDA(x); \ 25 | CHECK_CONTIGUOUS(x) 26 | 27 | int masked_im2col_forward_cuda(const at::Tensor im, const at::Tensor mask_h_idx, 28 | const at::Tensor mask_w_idx, const int kernel_h, 29 | const int kernel_w, const int pad_h, 30 | const int pad_w, at::Tensor col) { 31 | CHECK_INPUT(im); 32 | CHECK_INPUT(mask_h_idx); 33 | CHECK_INPUT(mask_w_idx); 34 | CHECK_INPUT(col); 35 | // im: (n, ic, h, w), kernel size (kh, kw) 36 | // kernel: (oc, ic * kh * kw), col: (kh * kw * ic, ow * oh) 37 | 38 | int channels = im.size(1); 39 | int height = im.size(2); 40 | int width = im.size(3); 41 | int mask_cnt = mask_h_idx.size(0); 42 | 43 | MaskedIm2colForwardLaucher(im, height, width, channels, kernel_h, kernel_w, 44 | pad_h, pad_w, mask_h_idx, mask_w_idx, mask_cnt, 45 | col); 46 | 47 | return 1; 48 | } 49 | 50 | int masked_col2im_forward_cuda(const at::Tensor col, 51 | const at::Tensor mask_h_idx, 52 | const at::Tensor mask_w_idx, int height, 53 | int width, int channels, at::Tensor im) { 54 | CHECK_INPUT(col); 55 | CHECK_INPUT(mask_h_idx); 56 | CHECK_INPUT(mask_w_idx); 57 | CHECK_INPUT(im); 58 | // im: (n, ic, h, w), kernel size (kh, kw) 59 | // kernel: (oc, ic * kh * kh), col: (kh * kw * ic, ow * oh) 60 | 61 | int mask_cnt = mask_h_idx.size(0); 62 | 63 | MaskedCol2imForwardLaucher(col, height, width, channels, mask_h_idx, 64 | mask_w_idx, mask_cnt, im); 65 | 66 | return 1; 67 | } 68 | 69 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { 70 | m.def("masked_im2col_forward", &masked_im2col_forward_cuda, 71 | "masked_im2col forward (CUDA)"); 72 | m.def("masked_col2im_forward", &masked_col2im_forward_cuda, 73 | "masked_col2im forward (CUDA)"); 74 | } -------------------------------------------------------------------------------- /mmdet/ops/ml_nms_rotated/__init__.py: -------------------------------------------------------------------------------- 1 | from .ml_nms_rotated_cuda import ml_nms_rotated 2 | 3 | __all__=['ml_nms_rotated'] 4 | -------------------------------------------------------------------------------- /mmdet/ops/ml_nms_rotated/src/nms_rotated.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved 2 | #pragma once 3 | #include 4 | #include 5 | 6 | at::Tensor nms_rotated_cpu( 7 | const at::Tensor& dets, 8 | const at::Tensor& scores, 9 | const at::Tensor& labels, 10 | const float iou_threshold); 11 | 12 | #ifdef WITH_CUDA 13 | at::Tensor nms_rotated_cuda( 14 | const at::Tensor& dets, 15 | const at::Tensor& scores, 16 | const at::Tensor& labels, 17 | const float iou_threshold); 18 | #endif 19 | 20 | // Interface for Python 21 | // inline is needed to prevent multiple function definitions when this header is 22 | // included by different cpps 23 | inline at::Tensor nms_rotated( 24 | const at::Tensor& dets, 25 | const at::Tensor& scores, 26 | const at::Tensor& labels, 27 | const float iou_threshold) { 28 | assert(dets.device().is_cuda() == scores.device().is_cuda()); 29 | if (dets.device().is_cuda()) { 30 | #ifdef WITH_CUDA 31 | return nms_rotated_cuda(dets, scores, labels, iou_threshold); 32 | #else 33 | AT_ERROR("Not compiled with GPU support"); 34 | #endif 35 | } 36 | 37 | return nms_rotated_cpu(dets, scores, labels, iou_threshold); 38 | } 39 | 40 | 41 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { 42 | m.def("ml_nms_rotated", &nms_rotated, "multi label NMS for rotated boxes"); 43 | 44 | } 45 | -------------------------------------------------------------------------------- /mmdet/ops/ml_nms_rotated/src/nms_rotated_cpu.cpp: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved 2 | #include "box_iou_rotated_utils.h" 3 | #include "nms_rotated.h" 4 | 5 | 6 | template 7 | at::Tensor nms_rotated_cpu_kernel( 8 | const at::Tensor& dets, 9 | const at::Tensor& scores, 10 | const float iou_threshold) { 11 | // nms_rotated_cpu_kernel is modified from torchvision's nms_cpu_kernel, 12 | // however, the code in this function is much shorter because 13 | // we delegate the IoU computation for rotated boxes to 14 | // the single_box_iou_rotated function in box_iou_rotated_utils.h 15 | AT_ASSERTM(!dets.type().is_cuda(), "dets must be a CPU tensor"); 16 | AT_ASSERTM(!scores.type().is_cuda(), "scores must be a CPU tensor"); 17 | AT_ASSERTM( 18 | dets.type() == scores.type(), "dets should have the same type as scores"); 19 | 20 | if (dets.numel() == 0) { 21 | return at::empty({0}, dets.options().dtype(at::kLong)); 22 | } 23 | 24 | auto order_t = std::get<1>(scores.sort(0, /* descending=*/true)); 25 | 26 | auto ndets = dets.size(0); 27 | at::Tensor suppressed_t = at::zeros({ndets}, dets.options().dtype(at::kByte)); 28 | at::Tensor keep_t = at::zeros({ndets}, dets.options().dtype(at::kLong)); 29 | 30 | auto suppressed = suppressed_t.data_ptr(); 31 | auto keep = keep_t.data_ptr(); 32 | auto order = order_t.data_ptr(); 33 | 34 | int64_t num_to_keep = 0; 35 | 36 | for (int64_t _i = 0; _i < ndets; _i++) { 37 | auto i = order[_i]; 38 | if (suppressed[i] == 1) { 39 | continue; 40 | } 41 | 42 | keep[num_to_keep++] = i; 43 | 44 | for (int64_t _j = _i + 1; _j < ndets; _j++) { 45 | auto j = order[_j]; 46 | if (suppressed[j] == 1) { 47 | continue; 48 | } 49 | 50 | auto ovr = single_box_iou_rotated( 51 | dets[i].data_ptr(), dets[j].data_ptr()); 52 | if (ovr >= iou_threshold) { 53 | suppressed[j] = 1; 54 | } 55 | } 56 | } 57 | return keep_t.narrow(/*dim=*/0, /*start=*/0, /*length=*/num_to_keep); 58 | } 59 | 60 | at::Tensor nms_rotated_cpu( 61 | const at::Tensor& dets, 62 | const at::Tensor& scores, 63 | const at::Tensor& labels, 64 | const float iou_threshold) { 65 | auto result = at::empty({0}, dets.options()); 66 | auto dets_wl = at::cat({dets, labels.unsqueeze(1)}, 1); 67 | AT_DISPATCH_FLOATING_TYPES(dets.type(), "nms_rotated", [&] { 68 | result = nms_rotated_cpu_kernel(dets_wl, scores, iou_threshold); 69 | }); 70 | return result; 71 | } 72 | 73 | -------------------------------------------------------------------------------- /mmdet/ops/nms/__init__.py: -------------------------------------------------------------------------------- 1 | from .nms_wrapper import nms, soft_nms 2 | 3 | __all__ = ['nms', 'soft_nms'] 4 | -------------------------------------------------------------------------------- /mmdet/ops/nms/nms_wrapper.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import torch 3 | 4 | from . import nms_cpu, nms_cuda 5 | from .soft_nms_cpu import soft_nms_cpu 6 | 7 | 8 | def nms(dets, iou_thr, device_id=None): 9 | """Dispatch to either CPU or GPU NMS implementations. 10 | 11 | The input can be either a torch tensor or numpy array. GPU NMS will be used 12 | if the input is a gpu tensor or device_id is specified, otherwise CPU NMS 13 | will be used. The returned type will always be the same as inputs. 14 | 15 | Arguments: 16 | dets (torch.Tensor or np.ndarray): bboxes with scores. 17 | iou_thr (float): IoU threshold for NMS. 18 | device_id (int, optional): when `dets` is a numpy array, if `device_id` 19 | is None, then cpu nms is used, otherwise gpu_nms will be used. 20 | 21 | Returns: 22 | tuple: kept bboxes and indice, which is always the same data type as 23 | the input. 24 | """ 25 | # convert dets (tensor or numpy array) to tensor 26 | if isinstance(dets, torch.Tensor): 27 | is_numpy = False 28 | dets_th = dets 29 | elif isinstance(dets, np.ndarray): 30 | is_numpy = True 31 | device = 'cpu' if device_id is None else 'cuda:{}'.format(device_id) 32 | dets_th = torch.from_numpy(dets).to(device) 33 | else: 34 | raise TypeError( 35 | 'dets must be either a Tensor or numpy array, but got {}'.format( 36 | type(dets))) 37 | 38 | # execute cpu or cuda nms 39 | if dets_th.shape[0] == 0: 40 | inds = dets_th.new_zeros(0, dtype=torch.long) 41 | else: 42 | if dets_th.is_cuda: 43 | inds = nms_cuda.nms(dets_th, iou_thr) 44 | else: 45 | inds = nms_cpu.nms(dets_th, iou_thr) 46 | 47 | if is_numpy: 48 | inds = inds.cpu().numpy() 49 | return dets[inds, :], inds 50 | 51 | 52 | def soft_nms(dets, iou_thr, method='linear', sigma=0.5, min_score=1e-3): 53 | if isinstance(dets, torch.Tensor): 54 | is_tensor = True 55 | dets_np = dets.detach().cpu().numpy() 56 | elif isinstance(dets, np.ndarray): 57 | is_tensor = False 58 | dets_np = dets 59 | else: 60 | raise TypeError( 61 | 'dets must be either a Tensor or numpy array, but got {}'.format( 62 | type(dets))) 63 | 64 | method_codes = {'linear': 1, 'gaussian': 2} 65 | if method not in method_codes: 66 | raise ValueError('Invalid method for SoftNMS: {}'.format(method)) 67 | new_dets, inds = soft_nms_cpu( 68 | dets_np, 69 | iou_thr, 70 | method=method_codes[method], 71 | sigma=sigma, 72 | min_score=min_score) 73 | 74 | if is_tensor: 75 | return dets.new_tensor(new_dets), dets.new_tensor( 76 | inds, dtype=torch.long) 77 | else: 78 | return new_dets.astype(np.float32), inds.astype(np.int64) 79 | -------------------------------------------------------------------------------- /mmdet/ops/nms/src/nms_cpu.cpp: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #include 3 | 4 | template 5 | at::Tensor nms_cpu_kernel(const at::Tensor& dets, const float threshold) { 6 | AT_ASSERTM(!dets.type().is_cuda(), "dets must be a CPU tensor"); 7 | 8 | if (dets.numel() == 0) { 9 | return at::empty({0}, dets.options().dtype(at::kLong).device(at::kCPU)); 10 | } 11 | 12 | auto x1_t = dets.select(1, 0).contiguous(); 13 | auto y1_t = dets.select(1, 1).contiguous(); 14 | auto x2_t = dets.select(1, 2).contiguous(); 15 | auto y2_t = dets.select(1, 3).contiguous(); 16 | auto scores = dets.select(1, 4).contiguous(); 17 | 18 | at::Tensor areas_t = (x2_t - x1_t + 1) * (y2_t - y1_t + 1); 19 | 20 | auto order_t = std::get<1>(scores.sort(0, /* descending=*/true)); 21 | 22 | auto ndets = dets.size(0); 23 | at::Tensor suppressed_t = 24 | at::zeros({ndets}, dets.options().dtype(at::kByte).device(at::kCPU)); 25 | 26 | auto suppressed = suppressed_t.data(); 27 | auto order = order_t.data(); 28 | auto x1 = x1_t.data(); 29 | auto y1 = y1_t.data(); 30 | auto x2 = x2_t.data(); 31 | auto y2 = y2_t.data(); 32 | auto areas = areas_t.data(); 33 | 34 | for (int64_t _i = 0; _i < ndets; _i++) { 35 | auto i = order[_i]; 36 | if (suppressed[i] == 1) continue; 37 | auto ix1 = x1[i]; 38 | auto iy1 = y1[i]; 39 | auto ix2 = x2[i]; 40 | auto iy2 = y2[i]; 41 | auto iarea = areas[i]; 42 | 43 | for (int64_t _j = _i + 1; _j < ndets; _j++) { 44 | auto j = order[_j]; 45 | if (suppressed[j] == 1) continue; 46 | auto xx1 = std::max(ix1, x1[j]); 47 | auto yy1 = std::max(iy1, y1[j]); 48 | auto xx2 = std::min(ix2, x2[j]); 49 | auto yy2 = std::min(iy2, y2[j]); 50 | 51 | auto w = std::max(static_cast(0), xx2 - xx1 + 1); 52 | auto h = std::max(static_cast(0), yy2 - yy1 + 1); 53 | auto inter = w * h; 54 | auto ovr = inter / (iarea + areas[j] - inter); 55 | if (ovr >= threshold) suppressed[j] = 1; 56 | } 57 | } 58 | return at::nonzero(suppressed_t == 0).squeeze(1); 59 | } 60 | 61 | at::Tensor nms(const at::Tensor& dets, const float threshold) { 62 | at::Tensor result; 63 | AT_DISPATCH_FLOATING_TYPES(dets.scalar_type(), "nms", [&] { 64 | result = nms_cpu_kernel(dets, threshold); 65 | }); 66 | return result; 67 | } 68 | 69 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { 70 | m.def("nms", &nms, "non-maximum suppression"); 71 | } -------------------------------------------------------------------------------- /mmdet/ops/nms/src/nms_cuda.cpp: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #include 3 | 4 | #define CHECK_CUDA(x) AT_CHECK(x.type().is_cuda(), #x, " must be a CUDAtensor ") 5 | 6 | at::Tensor nms_cuda(const at::Tensor boxes, float nms_overlap_thresh); 7 | 8 | at::Tensor nms(const at::Tensor& dets, const float threshold) { 9 | CHECK_CUDA(dets); 10 | if (dets.numel() == 0) 11 | return at::empty({0}, dets.options().dtype(at::kLong).device(at::kCPU)); 12 | return nms_cuda(dets, threshold); 13 | } 14 | 15 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { 16 | m.def("nms", &nms, "non-maximum suppression"); 17 | } -------------------------------------------------------------------------------- /mmdet/ops/nms_rotated/__init__.py: -------------------------------------------------------------------------------- 1 | from . import nms_rotated_cuda 2 | 3 | __all__ = ['nms_rotated'] 4 | 5 | 6 | def nms_rotated(dets, iou_thr): 7 | if dets.shape[0] == 0: 8 | return dets 9 | keep_inds = nms_rotated_cuda.nms_rotated(dets[:, :5], dets[:, 5], iou_thr) 10 | dets = dets[keep_inds, :] 11 | return dets, keep_inds 12 | -------------------------------------------------------------------------------- /mmdet/ops/nms_rotated/src/nms_rotated.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved 2 | #pragma once 3 | #include 4 | #include 5 | 6 | at::Tensor nms_rotated_cpu( 7 | const at::Tensor& dets, 8 | const at::Tensor& scores, 9 | const float iou_threshold); 10 | 11 | #ifdef WITH_CUDA 12 | at::Tensor nms_rotated_cuda( 13 | const at::Tensor& dets, 14 | const at::Tensor& scores, 15 | const float iou_threshold); 16 | #endif 17 | 18 | // Interface for Python 19 | // inline is needed to prevent multiple function definitions when this header is 20 | // included by different cpps 21 | inline at::Tensor nms_rotated( 22 | const at::Tensor& dets, 23 | const at::Tensor& scores, 24 | const float iou_threshold) { 25 | assert(dets.device().is_cuda() == scores.device().is_cuda()); 26 | if (dets.device().is_cuda()) { 27 | #ifdef WITH_CUDA 28 | return nms_rotated_cuda(dets, scores, iou_threshold); 29 | #else 30 | AT_ERROR("Not compiled with GPU support"); 31 | #endif 32 | } 33 | 34 | return nms_rotated_cpu(dets, scores, iou_threshold); 35 | } 36 | 37 | 38 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { 39 | m.def("nms_rotated", &nms_rotated, "NMS for rotated boxes"); 40 | 41 | } 42 | -------------------------------------------------------------------------------- /mmdet/ops/nms_rotated/src/nms_rotated_cpu.cpp: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved 2 | #include "box_iou_rotated_utils.h" 3 | #include "nms_rotated.h" 4 | 5 | 6 | template 7 | at::Tensor nms_rotated_cpu_kernel( 8 | const at::Tensor& dets, 9 | const at::Tensor& scores, 10 | const float iou_threshold) { 11 | // nms_rotated_cpu_kernel is modified from torchvision's nms_cpu_kernel, 12 | // however, the code in this function is much shorter because 13 | // we delegate the IoU computation for rotated boxes to 14 | // the single_box_iou_rotated function in box_iou_rotated_utils.h 15 | AT_ASSERTM(!dets.type().is_cuda(), "dets must be a CPU tensor"); 16 | AT_ASSERTM(!scores.type().is_cuda(), "scores must be a CPU tensor"); 17 | AT_ASSERTM( 18 | dets.type() == scores.type(), "dets should have the same type as scores"); 19 | 20 | if (dets.numel() == 0) { 21 | return at::empty({0}, dets.options().dtype(at::kLong)); 22 | } 23 | 24 | auto order_t = std::get<1>(scores.sort(0, /* descending=*/true)); 25 | 26 | auto ndets = dets.size(0); 27 | at::Tensor suppressed_t = at::zeros({ndets}, dets.options().dtype(at::kByte)); 28 | at::Tensor keep_t = at::zeros({ndets}, dets.options().dtype(at::kLong)); 29 | 30 | auto suppressed = suppressed_t.data_ptr(); 31 | auto keep = keep_t.data_ptr(); 32 | auto order = order_t.data_ptr(); 33 | 34 | int64_t num_to_keep = 0; 35 | 36 | for (int64_t _i = 0; _i < ndets; _i++) { 37 | auto i = order[_i]; 38 | if (suppressed[i] == 1) { 39 | continue; 40 | } 41 | 42 | keep[num_to_keep++] = i; 43 | 44 | for (int64_t _j = _i + 1; _j < ndets; _j++) { 45 | auto j = order[_j]; 46 | if (suppressed[j] == 1) { 47 | continue; 48 | } 49 | 50 | auto ovr = single_box_iou_rotated( 51 | dets[i].data_ptr(), dets[j].data_ptr()); 52 | if (ovr >= iou_threshold) { 53 | suppressed[j] = 1; 54 | } 55 | } 56 | } 57 | return keep_t.narrow(/*dim=*/0, /*start=*/0, /*length=*/num_to_keep); 58 | } 59 | 60 | at::Tensor nms_rotated_cpu( 61 | const at::Tensor& dets, 62 | const at::Tensor& scores, 63 | const float iou_threshold) { 64 | auto result = at::empty({0}, dets.options()); 65 | 66 | AT_DISPATCH_FLOATING_TYPES(dets.type(), "nms_rotated", [&] { 67 | result = nms_rotated_cpu_kernel(dets, scores, iou_threshold); 68 | }); 69 | return result; 70 | } 71 | 72 | -------------------------------------------------------------------------------- /mmdet/ops/orn/__init__.py: -------------------------------------------------------------------------------- 1 | from .modules.ORConv import ORConv2d 2 | from .functions import rotation_invariant_encoding,RotationInvariantPooling 3 | -------------------------------------------------------------------------------- /mmdet/ops/orn/functions/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | from .active_rotating_filter import active_rotating_filter 4 | from .active_rotating_filter import ActiveRotatingFilter 5 | from .rotation_invariant_encoding import rotation_invariant_encoding 6 | from .rotation_invariant_encoding import RotationInvariantEncoding 7 | from .rotation_invariant_pooling import RotationInvariantPooling 8 | 9 | __all__ = ['ActiveRotatingFilter', 'active_rotating_filter', 'rotation_invariant_encoding', 'RotationInvariantEncoding', 'RotationInvariantPooling'] -------------------------------------------------------------------------------- /mmdet/ops/orn/functions/active_rotating_filter.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | from torch import nn 4 | from torch.autograd import Function 5 | from torch.autograd.function import once_differentiable 6 | from torch.nn.modules.utils import _pair 7 | 8 | from .. import orn_cuda 9 | #import _C 10 | 11 | 12 | class _ActiveRotatingFilter(Function): 13 | @staticmethod 14 | def forward(ctx, input, indices): 15 | indices = indices.byte() 16 | ctx.input = input 17 | output = orn_cuda.arf_forward(input, indices) 18 | ctx.save_for_backward(indices) 19 | return output 20 | 21 | @staticmethod 22 | @once_differentiable 23 | def backward(ctx, grad_output): 24 | indices, = ctx.saved_tensors 25 | input = ctx.input 26 | grad_input = orn_cuda.arf_backward(indices, grad_output) 27 | return grad_input, None 28 | 29 | 30 | active_rotating_filter = _ActiveRotatingFilter.apply 31 | 32 | 33 | class ActiveRotatingFilter(nn.Module): 34 | def __init__(self, indices): 35 | super(ActiveRotatingFilter, self).__init__() 36 | self.indices = indices 37 | 38 | def forward(self, input): 39 | return active_rotating_filter(input, self.indices) 40 | 41 | 42 | if __name__ == "__main__": 43 | 44 | import math 45 | def get_indices(nOrientation, nRotation, kernel_size, mode='fast'): 46 | kernel_indices = { 47 | 1: { 48 | 0: (1,), 49 | 45: (1,), 50 | 90: (1,), 51 | 135: (1,), 52 | 180: (1,), 53 | 225: (1,), 54 | 270: (1,), 55 | 315: (1,) 56 | }, 57 | 3: { 58 | 0: (1,2,3,4,5,6,7,8,9), 59 | 45: (2,3,6,1,5,9,4,7,8), 60 | 90: (3,6,9,2,5,8,1,4,7), 61 | 135: (6,9,8,3,5,7,2,1,4), 62 | 180: (9,8,7,6,5,4,3,2,1), 63 | 225: (8,7,4,9,5,1,6,3,2), 64 | 270: (7,4,1,8,5,2,9,6,3), 65 | 315: (4,1,2,7,5,3,8,9,6) 66 | } 67 | } 68 | delta_orientation = 360 / nOrientation 69 | delta_rotation = 360 / nRotation 70 | kH, kW = kernel_size 71 | indices = torch.ByteTensor(nOrientation * kH * kW, nRotation) 72 | for i in range(0, nOrientation): 73 | for j in range(0, kH * kW): 74 | for k in range(0, nRotation): 75 | angle = delta_rotation * k 76 | layer = (i + math.floor(angle / delta_orientation)) % nOrientation 77 | kernel = kernel_indices[kW][angle][j] 78 | indices[i * kH * kW + j, k] = int(layer * kH * kW + kernel) 79 | return indices.view(nOrientation, kH, kW, nRotation) 80 | 81 | out_channels = 4 82 | in_channels = 2 83 | nOrientation = 8 84 | nRotation = 8 85 | kernel_size = 3 86 | input = torch.randn(out_channels, in_channels, nOrientation, kernel_size, kernel_size) 87 | input.requires_grad = True 88 | input = input.double() 89 | indices = get_indices(nOrientation, nRotation, (kernel_size, kernel_size)) 90 | input = input.cuda() 91 | indices = indices.cuda() 92 | output = active_rotating_filter(input, indices) 93 | print(output.size()) 94 | res = torch.autograd.gradcheck(active_rotating_filter, (input, indices), raise_exception=True) 95 | print(res) 96 | -------------------------------------------------------------------------------- /mmdet/ops/orn/functions/rotation_invariant_encoding.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | from torch import nn 4 | from torch.autograd import Function 5 | from torch.autograd.function import once_differentiable 6 | from torch.nn.modules.utils import _pair 7 | 8 | from .. import orn_cuda 9 | 10 | class _RotationInvariantEncoding(Function): 11 | @staticmethod 12 | def forward(ctx, input, nOrientation, return_direction=False): 13 | ctx.nOrientation = nOrientation 14 | ctx.return_direction = return_direction 15 | mainDirection, output = orn_cuda.rie_forward(input, nOrientation) 16 | if return_direction: 17 | ctx.save_for_backward(input, mainDirection) 18 | ctx.mark_non_differentiable(mainDirection) 19 | return output, mainDirection 20 | else: 21 | ctx.save_for_backward(input) 22 | ctx.mainDirection = mainDirection 23 | return output 24 | 25 | @staticmethod 26 | @once_differentiable 27 | def backward(ctx, grad_output): 28 | if ctx.return_direction: 29 | input, mainDirection = ctx.saved_tensors 30 | else: 31 | input, = ctx.saved_tensors 32 | mainDirection = ctx.mainDirection 33 | grad_input = orn_cuda.rie_backward(mainDirection, grad_output, ctx.nOrientation) 34 | return grad_input, None, None 35 | 36 | 37 | rotation_invariant_encoding = _RotationInvariantEncoding.apply 38 | 39 | 40 | class RotationInvariantEncoding(nn.Module): 41 | def __init__(self, nOrientation, return_direction=False): 42 | super(RotationInvariantEncoding, self).__init__() 43 | self.nOrientation = nOrientation 44 | self.return_direction = return_direction 45 | 46 | def forward(self, input): 47 | return rotation_invariant_encoding(input, self.nOrientation, self.return_direction) 48 | 49 | 50 | if __name__ == '__main__': 51 | nOrientation = 8 52 | input = torch.randn(2,8,1,1).double()#.cuda() 53 | input.requires_grad = True 54 | output = rotation_invariant_encoding(input, nOrientation) 55 | # check grad 56 | res = torch.autograd.gradcheck(rotation_invariant_encoding, (input, nOrientation), raise_exception=True) 57 | print(res) 58 | -------------------------------------------------------------------------------- /mmdet/ops/orn/functions/rotation_invariant_pooling.py: -------------------------------------------------------------------------------- 1 | import torch 2 | from torch import nn 3 | from torch.nn import functional as F 4 | 5 | 6 | class RotationInvariantPooling(nn.Module): 7 | def __init__(self, nInputPlane, nOrientation=8): 8 | super(RotationInvariantPooling, self).__init__() 9 | self.nInputPlane = nInputPlane 10 | self.nOrientation = nOrientation 11 | 12 | hiddent_dim = int(nInputPlane / nOrientation) 13 | self.conv = nn.Sequential( 14 | nn.Conv2d(hiddent_dim, nInputPlane, 1, 1), 15 | nn.BatchNorm2d(nInputPlane), 16 | ) 17 | 18 | def forward(self, x): 19 | # x: [N, c, 1, w] 20 | ## first, max_pooling along orientation. 21 | N, c, h, w = x.size() 22 | x = x.view(N, -1, self.nOrientation, h, w) 23 | x, _ = x.max(dim=2, keepdim=False) # [N, nInputPlane/nOrientation, 1, w] 24 | # MODIFIED 25 | # x = self.conv(x) # [N, nInputPlane, 1, w] 26 | return x 27 | 28 | 29 | if __name__ == '__main__': 30 | inst = RotationInvariantPooling(512, 8) 31 | input = torch.randn(8, 512, 1, 25) 32 | output = inst(input) 33 | print(output.size()) 34 | -------------------------------------------------------------------------------- /mmdet/ops/orn/modules/__init__.py: -------------------------------------------------------------------------------- 1 | from .ORConv import ORConv2d 2 | #from .ORConv_v2 import ORConv2d_v2 3 | 4 | #__all__ = ['ORConv2d', 'ORConv2d_v2'] 5 | __all__ = ['ORConv2d'] 6 | -------------------------------------------------------------------------------- /mmdet/ops/orn/src/ActiveRotatingFilter.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #pragma once 3 | 4 | #include "./cpu/vision.h" 5 | 6 | #ifdef WITH_CUDA 7 | #include "./cuda/vision.h" 8 | #endif 9 | 10 | // Interface for Python 11 | at::Tensor ARF_forward(const at::Tensor& weight, 12 | const at::Tensor& indices) { 13 | if (weight.type().is_cuda()) { 14 | #ifdef WITH_CUDA 15 | return ARF_forward_cuda(weight, indices); 16 | #else 17 | AT_ERROR("Not compiled with GPU support"); 18 | #endif 19 | } 20 | return ARF_forward_cpu(weight, indices); 21 | } 22 | 23 | at::Tensor ARF_backward(const at::Tensor& indices, 24 | const at::Tensor& gradOutput) { 25 | if (gradOutput.type().is_cuda()) { 26 | #ifdef WITH_CUDA 27 | return ARF_backward_cuda(indices, gradOutput); 28 | #else 29 | AT_ERROR("Not compiled with GPU support"); 30 | #endif 31 | } 32 | return ARF_backward_cpu(indices, gradOutput); 33 | AT_ERROR("Not implemented on the CPU"); 34 | } -------------------------------------------------------------------------------- /mmdet/ops/orn/src/RotationInvariantEncoding.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #pragma once 3 | 4 | #include "./cpu/vision.h" 5 | 6 | #ifdef WITH_CUDA 7 | #include "./cuda/vision.h" 8 | #endif 9 | 10 | // Interface for Python 11 | std::tuple RIE_forward(const at::Tensor& feature, 12 | const uint8 nOrientation) { 13 | if (feature.type().is_cuda()) { 14 | #ifdef WITH_CUDA 15 | return RIE_forward_cuda(feature, nOrientation); 16 | #else 17 | AT_ERROR("Not compiled with GPU support"); 18 | #endif 19 | } 20 | return RIE_forward_cpu(feature, nOrientation); 21 | } 22 | 23 | at::Tensor RIE_backward(const at::Tensor& mainDirection, 24 | const at::Tensor& gradOutput, 25 | const uint8 nOrientation) { 26 | if (gradOutput.type().is_cuda()) { 27 | #ifdef WITH_CUDA 28 | return RIE_backward_cuda(mainDirection, gradOutput, nOrientation); 29 | #else 30 | AT_ERROR("Not compiled with GPU support"); 31 | #endif 32 | } 33 | return RIE_backward_cpu(mainDirection, gradOutput, nOrientation); 34 | AT_ERROR("Not implemented on the CPU"); 35 | } -------------------------------------------------------------------------------- /mmdet/ops/orn/src/cpu/vision.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #pragma once 3 | // #include 4 | #include 5 | 6 | typedef unsigned long uint64; 7 | typedef unsigned int uint32; 8 | typedef unsigned short uint16; 9 | typedef unsigned char uint8; 10 | 11 | 12 | std::tuple RIE_forward_cpu(const at::Tensor& feature, 13 | const uint8 nOrientation); 14 | 15 | at::Tensor RIE_backward_cpu(const at::Tensor& mainDirection, 16 | const at::Tensor& gradOutput, 17 | const uint8 nOrientation); 18 | 19 | at::Tensor ARF_forward_cpu(const at::Tensor& weight, 20 | const at::Tensor& indices); 21 | 22 | at::Tensor ARF_backward_cpu(const at::Tensor& indices, 23 | const at::Tensor& gradOutput); -------------------------------------------------------------------------------- /mmdet/ops/orn/src/cuda/vision.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #pragma once 3 | // #include 4 | #include 5 | 6 | typedef unsigned long uint64; 7 | typedef unsigned int uint32; 8 | typedef unsigned short uint16; 9 | typedef unsigned char uint8; 10 | 11 | 12 | std::tuple RIE_forward_cuda(const at::Tensor& feature, 13 | const uint8 nOrientation); 14 | 15 | at::Tensor RIE_backward_cuda(const at::Tensor& mainDirection, 16 | const at::Tensor& gradOutput, 17 | const uint8 nOrientation); 18 | 19 | at::Tensor ARF_forward_cuda(const at::Tensor& weight, 20 | const at::Tensor& indices); 21 | 22 | at::Tensor ARF_backward_cuda(const at::Tensor& indices, 23 | const at::Tensor& gradOutput); -------------------------------------------------------------------------------- /mmdet/ops/orn/src/vision.cpp: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #include 3 | #include "./ActiveRotatingFilter.h" 4 | #include "./RotationInvariantEncoding.h" 5 | 6 | 7 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { 8 | m.def("arf_forward", &ARF_forward, "ARF_forward"); 9 | m.def("arf_backward", &ARF_backward, "ARF_backward"); 10 | m.def("rie_forward", &RIE_forward, "RIE_forward"); 11 | m.def("rie_backward", &RIE_backward, "RIE_backward"); 12 | } -------------------------------------------------------------------------------- /mmdet/ops/roi_align/__init__.py: -------------------------------------------------------------------------------- 1 | from .roi_align import RoIAlign, roi_align 2 | 3 | __all__ = ['roi_align', 'RoIAlign'] 4 | -------------------------------------------------------------------------------- /mmdet/ops/roi_align/gradcheck.py: -------------------------------------------------------------------------------- 1 | import os.path as osp 2 | import sys 3 | 4 | import numpy as np 5 | import torch 6 | from torch.autograd import gradcheck 7 | 8 | sys.path.append(osp.abspath(osp.join(__file__, '../../'))) 9 | from roi_align import RoIAlign # noqa: E402, isort:skip 10 | 11 | feat_size = 15 12 | spatial_scale = 1.0 / 8 13 | img_size = feat_size / spatial_scale 14 | num_imgs = 2 15 | num_rois = 20 16 | 17 | batch_ind = np.random.randint(num_imgs, size=(num_rois, 1)) 18 | rois = np.random.rand(num_rois, 4) * img_size * 0.5 19 | rois[:, 2:] += img_size * 0.5 20 | rois = np.hstack((batch_ind, rois)) 21 | 22 | feat = torch.randn( 23 | num_imgs, 16, feat_size, feat_size, requires_grad=True, device='cuda:0') 24 | rois = torch.from_numpy(rois).float().cuda() 25 | inputs = (feat, rois) 26 | print('Gradcheck for roi align...') 27 | test = gradcheck(RoIAlign(3, spatial_scale), inputs, atol=1e-3, eps=1e-3) 28 | print(test) 29 | test = gradcheck(RoIAlign(3, spatial_scale, 2), inputs, atol=1e-3, eps=1e-3) 30 | print(test) 31 | -------------------------------------------------------------------------------- /mmdet/ops/roi_align_rotated/__init__.py: -------------------------------------------------------------------------------- 1 | from .roi_align_rotated import RoIAlignRotated 2 | 3 | __all__ = ['RoIAlignRotated'] 4 | -------------------------------------------------------------------------------- /mmdet/ops/roi_pool/__init__.py: -------------------------------------------------------------------------------- 1 | from .roi_pool import RoIPool, roi_pool 2 | 3 | __all__ = ['roi_pool', 'RoIPool'] 4 | -------------------------------------------------------------------------------- /mmdet/ops/roi_pool/gradcheck.py: -------------------------------------------------------------------------------- 1 | import os.path as osp 2 | import sys 3 | 4 | import torch 5 | from torch.autograd import gradcheck 6 | 7 | sys.path.append(osp.abspath(osp.join(__file__, '../../'))) 8 | from roi_pool import RoIPool # noqa: E402, isort:skip 9 | 10 | feat = torch.randn(4, 16, 15, 15, requires_grad=True).cuda() 11 | rois = torch.Tensor([[0, 0, 0, 50, 50], [0, 10, 30, 43, 55], 12 | [1, 67, 40, 110, 120]]).cuda() 13 | inputs = (feat, rois) 14 | print('Gradcheck for roi pooling...') 15 | test = gradcheck(RoIPool(4, 1.0 / 8), inputs, eps=1e-5, atol=1e-3) 16 | print(test) 17 | -------------------------------------------------------------------------------- /mmdet/ops/roi_pool/roi_pool.py: -------------------------------------------------------------------------------- 1 | import torch 2 | import torch.nn as nn 3 | from torch.autograd import Function 4 | from torch.autograd.function import once_differentiable 5 | from torch.nn.modules.utils import _pair 6 | 7 | from . import roi_pool_cuda 8 | 9 | 10 | class RoIPoolFunction(Function): 11 | 12 | @staticmethod 13 | def forward(ctx, features, rois, out_size, spatial_scale): 14 | assert features.is_cuda 15 | out_h, out_w = _pair(out_size) 16 | assert isinstance(out_h, int) and isinstance(out_w, int) 17 | ctx.save_for_backward(rois) 18 | num_channels = features.size(1) 19 | num_rois = rois.size(0) 20 | out_size = (num_rois, num_channels, out_h, out_w) 21 | output = features.new_zeros(out_size) 22 | argmax = features.new_zeros(out_size, dtype=torch.int) 23 | roi_pool_cuda.forward(features, rois, out_h, out_w, spatial_scale, 24 | output, argmax) 25 | ctx.spatial_scale = spatial_scale 26 | ctx.feature_size = features.size() 27 | ctx.argmax = argmax 28 | 29 | return output 30 | 31 | @staticmethod 32 | @once_differentiable 33 | def backward(ctx, grad_output): 34 | assert grad_output.is_cuda 35 | spatial_scale = ctx.spatial_scale 36 | feature_size = ctx.feature_size 37 | argmax = ctx.argmax 38 | rois = ctx.saved_tensors[0] 39 | assert feature_size is not None 40 | 41 | grad_input = grad_rois = None 42 | if ctx.needs_input_grad[0]: 43 | grad_input = grad_output.new_zeros(feature_size) 44 | roi_pool_cuda.backward(grad_output.contiguous(), rois, argmax, 45 | spatial_scale, grad_input) 46 | 47 | return grad_input, grad_rois, None, None 48 | 49 | 50 | roi_pool = RoIPoolFunction.apply 51 | 52 | 53 | class RoIPool(nn.Module): 54 | 55 | def __init__(self, out_size, spatial_scale, use_torchvision=False): 56 | super(RoIPool, self).__init__() 57 | 58 | self.out_size = _pair(out_size) 59 | self.spatial_scale = float(spatial_scale) 60 | self.use_torchvision = use_torchvision 61 | 62 | def forward(self, features, rois): 63 | if self.use_torchvision: 64 | from torchvision.ops import roi_pool as tv_roi_pool 65 | return tv_roi_pool(features, rois, self.out_size, 66 | self.spatial_scale) 67 | else: 68 | return roi_pool(features, rois, self.out_size, self.spatial_scale) 69 | 70 | def __repr__(self): 71 | format_str = self.__class__.__name__ 72 | format_str += '(out_size={}, spatial_scale={}'.format( 73 | self.out_size, self.spatial_scale) 74 | format_str += ', use_torchvision={})'.format(self.use_torchvision) 75 | return format_str 76 | -------------------------------------------------------------------------------- /mmdet/ops/sigmoid_focal_loss/__init__.py: -------------------------------------------------------------------------------- 1 | from .sigmoid_focal_loss import SigmoidFocalLoss, sigmoid_focal_loss 2 | 3 | __all__ = ['SigmoidFocalLoss', 'sigmoid_focal_loss'] 4 | -------------------------------------------------------------------------------- /mmdet/ops/sigmoid_focal_loss/sigmoid_focal_loss.py: -------------------------------------------------------------------------------- 1 | import torch.nn as nn 2 | from torch.autograd import Function 3 | from torch.autograd.function import once_differentiable 4 | 5 | from . import sigmoid_focal_loss_cuda 6 | 7 | 8 | class SigmoidFocalLossFunction(Function): 9 | 10 | @staticmethod 11 | def forward(ctx, input, target, gamma=2.0, alpha=0.25): 12 | ctx.save_for_backward(input, target) 13 | num_classes = input.shape[1] 14 | ctx.num_classes = num_classes 15 | ctx.gamma = gamma 16 | ctx.alpha = alpha 17 | 18 | loss = sigmoid_focal_loss_cuda.forward(input, target, num_classes, 19 | gamma, alpha) 20 | return loss 21 | 22 | @staticmethod 23 | @once_differentiable 24 | def backward(ctx, d_loss): 25 | input, target = ctx.saved_tensors 26 | num_classes = ctx.num_classes 27 | gamma = ctx.gamma 28 | alpha = ctx.alpha 29 | d_loss = d_loss.contiguous() 30 | d_input = sigmoid_focal_loss_cuda.backward(input, target, d_loss, 31 | num_classes, gamma, alpha) 32 | return d_input, None, None, None, None 33 | 34 | 35 | sigmoid_focal_loss = SigmoidFocalLossFunction.apply 36 | 37 | 38 | # TODO: remove this module 39 | class SigmoidFocalLoss(nn.Module): 40 | 41 | def __init__(self, gamma, alpha): 42 | super(SigmoidFocalLoss, self).__init__() 43 | self.gamma = gamma 44 | self.alpha = alpha 45 | 46 | def forward(self, logits, targets): 47 | assert logits.is_cuda 48 | loss = sigmoid_focal_loss(logits, targets, self.gamma, self.alpha) 49 | return loss.sum() 50 | 51 | def __repr__(self): 52 | tmpstr = self.__class__.__name__ + '(gamma={}, alpha={})'.format( 53 | self.gamma, self.alpha) 54 | return tmpstr 55 | -------------------------------------------------------------------------------- /mmdet/ops/sigmoid_focal_loss/src/sigmoid_focal_loss.cpp: -------------------------------------------------------------------------------- 1 | // modify from 2 | // https://github.com/facebookresearch/maskrcnn-benchmark/blob/master/maskrcnn_benchmark/csrc/SigmoidFocalLoss.h 3 | #include 4 | 5 | at::Tensor SigmoidFocalLoss_forward_cuda(const at::Tensor &logits, 6 | const at::Tensor &targets, 7 | const int num_classes, 8 | const float gamma, const float alpha); 9 | 10 | at::Tensor SigmoidFocalLoss_backward_cuda(const at::Tensor &logits, 11 | const at::Tensor &targets, 12 | const at::Tensor &d_losses, 13 | const int num_classes, 14 | const float gamma, const float alpha); 15 | 16 | // Interface for Python 17 | at::Tensor SigmoidFocalLoss_forward(const at::Tensor &logits, 18 | const at::Tensor &targets, 19 | const int num_classes, const float gamma, 20 | const float alpha) { 21 | if (logits.type().is_cuda()) { 22 | return SigmoidFocalLoss_forward_cuda(logits, targets, num_classes, gamma, 23 | alpha); 24 | } 25 | AT_ERROR("SigmoidFocalLoss is not implemented on the CPU"); 26 | } 27 | 28 | at::Tensor SigmoidFocalLoss_backward(const at::Tensor &logits, 29 | const at::Tensor &targets, 30 | const at::Tensor &d_losses, 31 | const int num_classes, const float gamma, 32 | const float alpha) { 33 | if (logits.type().is_cuda()) { 34 | return SigmoidFocalLoss_backward_cuda(logits, targets, d_losses, 35 | num_classes, gamma, alpha); 36 | } 37 | AT_ERROR("SigmoidFocalLoss is not implemented on the CPU"); 38 | } 39 | 40 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { 41 | m.def("forward", &SigmoidFocalLoss_forward, 42 | "SigmoidFocalLoss forward (CUDA)"); 43 | m.def("backward", &SigmoidFocalLoss_backward, 44 | "SigmoidFocalLoss backward (CUDA)"); 45 | } 46 | -------------------------------------------------------------------------------- /mmdet/utils/__init__.py: -------------------------------------------------------------------------------- 1 | from .flops_counter import get_model_complexity_info 2 | from .registry import Registry, build_from_cfg 3 | 4 | __all__ = ['Registry', 'build_from_cfg', 'get_model_complexity_info'] 5 | -------------------------------------------------------------------------------- /mmdet/utils/registry.py: -------------------------------------------------------------------------------- 1 | import inspect 2 | 3 | import mmcv 4 | 5 | 6 | class Registry(object): 7 | 8 | def __init__(self, name): 9 | self._name = name 10 | self._module_dict = dict() 11 | 12 | def __repr__(self): 13 | format_str = self.__class__.__name__ + '(name={}, items={})'.format( 14 | self._name, list(self._module_dict.keys())) 15 | return format_str 16 | 17 | @property 18 | def name(self): 19 | return self._name 20 | 21 | @property 22 | def module_dict(self): 23 | return self._module_dict 24 | 25 | def get(self, key): 26 | return self._module_dict.get(key, None) 27 | 28 | def _register_module(self, module_class): 29 | """Register a module. 30 | 31 | Args: 32 | module (:obj:`nn.Module`): Module to be registered. 33 | """ 34 | if not inspect.isclass(module_class): 35 | raise TypeError('module must be a class, but got {}'.format( 36 | type(module_class))) 37 | module_name = module_class.__name__ 38 | if module_name in self._module_dict: 39 | raise KeyError('{} is already registered in {}'.format( 40 | module_name, self.name)) 41 | self._module_dict[module_name] = module_class 42 | 43 | def register_module(self, cls): 44 | self._register_module(cls) 45 | return cls 46 | 47 | 48 | def build_from_cfg(cfg, registry, default_args=None): 49 | """Build a module from config dict. 50 | 51 | Args: 52 | cfg (dict): Config dict. It should at least contain the key "type". 53 | registry (:obj:`Registry`): The registry to search the type from. 54 | default_args (dict, optional): Default initialization arguments. 55 | 56 | Returns: 57 | obj: The constructed object. 58 | """ 59 | assert isinstance(cfg, dict) and 'type' in cfg 60 | assert isinstance(default_args, dict) or default_args is None 61 | args = cfg.copy() 62 | obj_type = args.pop('type') 63 | if mmcv.is_str(obj_type): 64 | obj_cls = registry.get(obj_type) 65 | if obj_cls is None: 66 | raise KeyError('{} is not in the {} registry'.format( 67 | obj_type, registry.name)) 68 | elif inspect.isclass(obj_type): 69 | obj_cls = obj_type 70 | else: 71 | raise TypeError('type must be a str or valid type, but got {}'.format( 72 | type(obj_type))) 73 | if default_args is not None: 74 | for name, value in default_args.items(): 75 | args.setdefault(name, value) 76 | return obj_cls(**args) 77 | -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | cython 2 | numpy 3 | albumentations 4 | imagecorruptions 5 | mmcv==0.2.14 6 | matplotlib 7 | terminaltables 8 | six 9 | pycocotools 10 | pillow==6.2.2 11 | shapely -------------------------------------------------------------------------------- /tools/coco_eval.py: -------------------------------------------------------------------------------- 1 | from argparse import ArgumentParser 2 | 3 | from mmdet.core import coco_eval 4 | 5 | 6 | def main(): 7 | parser = ArgumentParser(description='COCO Evaluation') 8 | parser.add_argument('result', help='result file path') 9 | parser.add_argument('--ann', help='annotation file path') 10 | parser.add_argument( 11 | '--types', 12 | type=str, 13 | nargs='+', 14 | choices=['proposal_fast', 'proposal', 'bbox', 'segm', 'keypoint'], 15 | default=['bbox'], 16 | help='result types') 17 | parser.add_argument( 18 | '--max-dets', 19 | type=int, 20 | nargs='+', 21 | default=[100, 300, 1000], 22 | help='proposal numbers, only used for recall evaluation') 23 | parser.add_argument( 24 | '--classwise', action='store_true', help='whether eval class wise ap') 25 | args = parser.parse_args() 26 | coco_eval(args.result, args.types, args.ann, args.max_dets, args.classwise) 27 | 28 | 29 | if __name__ == '__main__': 30 | main() 31 | -------------------------------------------------------------------------------- /tools/convert_model.py: -------------------------------------------------------------------------------- 1 | import argparse 2 | import subprocess 3 | from collections import OrderedDict 4 | import torch 5 | 6 | 7 | def parse_args(): 8 | parser = argparse.ArgumentParser( 9 | description='Process a checkpoint to be published') 10 | parser.add_argument('in_file', help='input checkpoint filename') 11 | parser.add_argument('out_file', help='output checkpoint filename') 12 | args = parser.parse_args() 13 | return args 14 | 15 | 16 | def process_checkpoint(in_file, out_file): 17 | checkpoint = torch.load(in_file, map_location='cpu') 18 | # remove optimizer for smaller file size 19 | if 'optimizer' in checkpoint: 20 | del checkpoint['optimizer'] 21 | in_state_dict = checkpoint.pop('state_dict') 22 | out_state_dict = OrderedDict() 23 | for key, val in in_state_dict.items(): 24 | if 'rbox_head' in key: 25 | key = key.replace('rbox_head','bbox_head') 26 | out_state_dict[key] = val 27 | checkpoint['state_dict'] = out_state_dict 28 | 29 | # if it is necessary to remove some sensitive data in checkpoint['meta'], 30 | # add the code here. 31 | torch.save(checkpoint, out_file) 32 | sha = subprocess.check_output(['sha256sum', out_file]).decode() 33 | final_file = out_file.rstrip('.pth') + '-{}.pth'.format(sha[:8]) 34 | subprocess.Popen(['mv', out_file, final_file]) 35 | 36 | 37 | def main(): 38 | args = parse_args() 39 | process_checkpoint(args.in_file, args.out_file) 40 | 41 | 42 | if __name__ == '__main__': 43 | main() 44 | -------------------------------------------------------------------------------- /tools/dist_test.sh: -------------------------------------------------------------------------------- 1 | #!/usr/bin/env bash 2 | 3 | PYTHON=${PYTHON:-"python"} 4 | 5 | CONFIG=$1 6 | CHECKPOINT=$2 7 | GPUS=$3 8 | 9 | $PYTHON -m torch.distributed.launch --nproc_per_node=$GPUS \ 10 | $(dirname "$0")/test.py $CONFIG $CHECKPOINT --launcher pytorch ${@:4} 11 | -------------------------------------------------------------------------------- /tools/dist_train.sh: -------------------------------------------------------------------------------- 1 | #!/usr/bin/env bash 2 | 3 | PYTHON=${PYTHON:-"python"} 4 | 5 | CONFIG=$1 6 | GPUS=$2 7 | 8 | $PYTHON -m torch.distributed.launch --nproc_per_node=$GPUS \ 9 | $(dirname "$0")/train.py $CONFIG --launcher pytorch ${@:3} 10 | -------------------------------------------------------------------------------- /tools/get_flops.py: -------------------------------------------------------------------------------- 1 | import argparse 2 | 3 | from mmcv import Config 4 | 5 | from mmdet.models import build_detector 6 | from mmdet.utils import get_model_complexity_info 7 | 8 | 9 | def parse_args(): 10 | parser = argparse.ArgumentParser(description='Train a detector') 11 | parser.add_argument('config', help='train config file path') 12 | parser.add_argument( 13 | '--shape', 14 | type=int, 15 | nargs='+', 16 | default=[1280, 800], 17 | help='input image size') 18 | args = parser.parse_args() 19 | return args 20 | 21 | 22 | def main(): 23 | 24 | args = parse_args() 25 | 26 | if len(args.shape) == 1: 27 | input_shape = (3, args.shape[0], args.shape[0]) 28 | elif len(args.shape) == 2: 29 | input_shape = (3, ) + tuple(args.shape) 30 | else: 31 | raise ValueError('invalid input shape') 32 | 33 | cfg = Config.fromfile(args.config) 34 | model = build_detector( 35 | cfg.model, train_cfg=cfg.train_cfg, test_cfg=cfg.test_cfg).cuda() 36 | model.eval() 37 | 38 | if hasattr(model, 'forward_dummy'): 39 | model.forward = model.forward_dummy 40 | else: 41 | raise NotImplementedError( 42 | 'FLOPs counter is currently not currently supported with {}'. 43 | format(model.__class__.__name__)) 44 | 45 | flops, params = get_model_complexity_info(model, input_shape) 46 | split_line = '=' * 30 47 | print('{0}\nInput shape: {1}\nFlops: {2}\nParams: {3}\n{0}'.format( 48 | split_line, input_shape, flops, params)) 49 | 50 | 51 | if __name__ == '__main__': 52 | main() 53 | -------------------------------------------------------------------------------- /tools/publish_model.py: -------------------------------------------------------------------------------- 1 | import argparse 2 | import subprocess 3 | 4 | import torch 5 | 6 | 7 | def parse_args(): 8 | parser = argparse.ArgumentParser( 9 | description='Process a checkpoint to be published') 10 | parser.add_argument('in_file', help='input checkpoint filename') 11 | parser.add_argument('out_file', help='output checkpoint filename') 12 | args = parser.parse_args() 13 | return args 14 | 15 | 16 | def process_checkpoint(in_file, out_file): 17 | checkpoint = torch.load(in_file, map_location='cpu') 18 | # remove optimizer for smaller file size 19 | if 'optimizer' in checkpoint: 20 | del checkpoint['optimizer'] 21 | # if it is necessary to remove some sensitive data in checkpoint['meta'], 22 | # add the code here. 23 | torch.save(checkpoint, out_file) 24 | sha = subprocess.check_output(['sha256sum', out_file]).decode() 25 | final_file = out_file.rstrip('.pth') + '-{}.pth'.format(sha[:8]) 26 | subprocess.Popen(['mv', out_file, final_file]) 27 | 28 | 29 | def main(): 30 | args = parse_args() 31 | process_checkpoint(args.in_file, args.out_file) 32 | 33 | 34 | if __name__ == '__main__': 35 | main() 36 | -------------------------------------------------------------------------------- /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:-4} 10 | GPUS_PER_NODE=${GPUS_PER_NODE:-4} 11 | CPUS_PER_TASK=${CPUS_PER_TASK:-4} 12 | PY_ARGS=${@:5} 13 | SRUN_ARGS=${SRUN_ARGS:-""} 14 | 15 | srun -p ${PARTITION} \ 16 | --job-name=${JOB_NAME} \ 17 | --gres=gpu:${GPUS_PER_NODE} \ 18 | --ntasks=${GPUS} \ 19 | --ntasks-per-node=${GPUS_PER_NODE} \ 20 | --cpus-per-task=${CPUS_PER_TASK} \ 21 | --kill-on-bad-exit=1 \ 22 | ${SRUN_ARGS} \ 23 | python -u tools/test.py ${CONFIG} ${CHECKPOINT} --launcher="slurm" ${PY_ARGS} 24 | -------------------------------------------------------------------------------- /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=$4 9 | GPUS=${5:-4} 10 | GPUS_PER_NODE=${GPUS_PER_NODE:-4} 11 | CPUS_PER_TASK=${CPUS_PER_TASK:-4} 12 | SRUN_ARGS=${SRUN_ARGS:-""} 13 | # PY_ARGS=${PY_ARGS:-"--validate"} 14 | PY_ARGS=${PY_ARGS:-""} 15 | 16 | 17 | srun -p ${PARTITION} \ 18 | --job-name=${JOB_NAME} \ 19 | --gres=gpu:${GPUS_PER_NODE} \ 20 | --ntasks=${GPUS} \ 21 | --ntasks-per-node=${GPUS_PER_NODE} \ 22 | --cpus-per-task=${CPUS_PER_TASK} \ 23 | --kill-on-bad-exit=1 \ 24 | ${SRUN_ARGS} \ 25 | python -u tools/train.py ${CONFIG} --work_dir=${WORK_DIR} --launcher="slurm" ${PY_ARGS} 26 | -------------------------------------------------------------------------------- /tools/slurm_train.slurm: -------------------------------------------------------------------------------- 1 | #!/usr/bin/env bash 2 | 3 | #SBATCH -p gpu 4 | #SBATCH --ntasks=4 5 | #SBATCH --ntasks-per-node=4 6 | #SBATCH --cpus-per-task=4 7 | #SBATCH --gres=gpu:4 8 | #SBATCH -o train_s2anet_r50_fpn_1x.log 9 | 10 | module load scl/gcc4.9 11 | module load nvidia/cuda/10.0 12 | nvidia-smi 13 | ./tools/dist_train.sh \ 14 | configs/dota/s2anet_r50_fpn_1x.py 4 -------------------------------------------------------------------------------- /tools/upgrade_model_version.py: -------------------------------------------------------------------------------- 1 | import argparse 2 | import re 3 | from collections import OrderedDict 4 | 5 | import torch 6 | 7 | 8 | def convert(in_file, out_file): 9 | """Convert keys in checkpoints. 10 | 11 | There can be some breaking changes during the development of mmdetection, 12 | and this tool is used for upgrading checkpoints trained with old versions 13 | to the latest one. 14 | """ 15 | checkpoint = torch.load(in_file) 16 | in_state_dict = checkpoint.pop('state_dict') 17 | out_state_dict = OrderedDict() 18 | for key, val in in_state_dict.items(): 19 | # Use ConvModule instead of nn.Conv2d in RetinaNet 20 | # cls_convs.0.weight -> cls_convs.0.conv.weight 21 | m = re.search(r'(cls_convs|reg_convs).\d.(weight|bias)', key) 22 | if m is not None: 23 | param = m.groups()[1] 24 | new_key = key.replace(param, 'conv.{}'.format(param)) 25 | out_state_dict[new_key] = val 26 | continue 27 | 28 | out_state_dict[key] = val 29 | checkpoint['state_dict'] = out_state_dict 30 | torch.save(checkpoint, out_file) 31 | 32 | 33 | def main(): 34 | parser = argparse.ArgumentParser(description='Upgrade model version') 35 | parser.add_argument('in_file', help='input checkpoint file') 36 | parser.add_argument('out_file', help='output checkpoint file') 37 | args = parser.parse_args() 38 | convert(args.in_file, args.out_file) 39 | 40 | 41 | if __name__ == '__main__': 42 | main() 43 | -------------------------------------------------------------------------------- /tools/voc_eval.py: -------------------------------------------------------------------------------- 1 | from argparse import ArgumentParser 2 | 3 | import mmcv 4 | import numpy as np 5 | 6 | from mmdet import datasets 7 | from mmdet.core import eval_map 8 | 9 | 10 | def voc_eval(result_file, dataset, iou_thr=0.5): 11 | det_results = mmcv.load(result_file) 12 | gt_bboxes = [] 13 | gt_labels = [] 14 | gt_ignore = [] 15 | for i in range(len(dataset)): 16 | ann = dataset.get_ann_info(i) 17 | bboxes = ann['bboxes'] 18 | labels = ann['labels'] 19 | if 'bboxes_ignore' in ann: 20 | ignore = np.concatenate([ 21 | np.zeros(bboxes.shape[0], dtype=np.bool), 22 | np.ones(ann['bboxes_ignore'].shape[0], dtype=np.bool) 23 | ]) 24 | gt_ignore.append(ignore) 25 | bboxes = np.vstack([bboxes, ann['bboxes_ignore']]) 26 | labels = np.concatenate([labels, ann['labels_ignore']]) 27 | gt_bboxes.append(bboxes) 28 | gt_labels.append(labels) 29 | if not gt_ignore: 30 | gt_ignore = None 31 | if hasattr(dataset, 'year') and dataset.year == 2007: 32 | dataset_name = 'voc07' 33 | else: 34 | dataset_name = dataset.CLASSES 35 | eval_map( 36 | det_results, 37 | gt_bboxes, 38 | gt_labels, 39 | gt_ignore=gt_ignore, 40 | scale_ranges=None, 41 | iou_thr=iou_thr, 42 | dataset=dataset_name, 43 | print_summary=True) 44 | 45 | 46 | def main(): 47 | parser = ArgumentParser(description='VOC Evaluation') 48 | parser.add_argument('result', help='result file path') 49 | parser.add_argument('config', help='config file path') 50 | parser.add_argument( 51 | '--iou-thr', 52 | type=float, 53 | default=0.5, 54 | help='IoU threshold for evaluation') 55 | args = parser.parse_args() 56 | cfg = mmcv.Config.fromfile(args.config) 57 | test_dataset = mmcv.runner.obj_from_dict(cfg.data.test, datasets) 58 | voc_eval(args.result, test_dataset, args.iou_thr) 59 | 60 | 61 | if __name__ == '__main__': 62 | main() 63 | --------------------------------------------------------------------------------