├── maskrcnn_benchmark ├── utils │ ├── __init__.py │ ├── __pycache__ │ │ ├── comm.cpython-36.pyc │ │ ├── __init__.cpython-36.pyc │ │ ├── cv2_util.cpython-36.pyc │ │ ├── imports.cpython-36.pyc │ │ ├── registry.cpython-36.pyc │ │ ├── checkpoint.cpython-36.pyc │ │ ├── model_zoo.cpython-36.pyc │ │ ├── miscellaneous.cpython-36.pyc │ │ ├── c2_model_loading.cpython-36.pyc │ │ └── model_serialization.cpython-36.pyc │ ├── README.md │ ├── collect_env.py │ ├── cv2_util.py │ ├── logger.py │ ├── imports.py │ ├── miscellaneous.py │ ├── timer.py │ ├── env.py │ ├── registry.py │ ├── metric_logger.py │ ├── model_zoo.py │ ├── model_serialization.py │ └── comm.py ├── modeling │ ├── __init__.py │ ├── roi_heads │ │ ├── __init__.py │ │ ├── box_head │ │ │ ├── __init__.py │ │ │ ├── __pycache__ │ │ │ │ ├── loss.cpython-36.pyc │ │ │ │ ├── __init__.cpython-36.pyc │ │ │ │ ├── box_head.cpython-36.pyc │ │ │ │ ├── inference.cpython-36.pyc │ │ │ │ ├── roi_box_predictors.cpython-36.pyc │ │ │ │ └── roi_box_feature_extractors.cpython-36.pyc │ │ │ ├── roi_box_predictors.py │ │ │ └── box_head.py │ │ ├── keypoint_head │ │ │ ├── __init__.py │ │ │ ├── __pycache__ │ │ │ │ ├── loss.cpython-36.pyc │ │ │ │ ├── __init__.cpython-36.pyc │ │ │ │ ├── inference.cpython-36.pyc │ │ │ │ ├── keypoint_head.cpython-36.pyc │ │ │ │ ├── roi_keypoint_predictors.cpython-36.pyc │ │ │ │ └── roi_keypoint_feature_extractors.cpython-36.pyc │ │ │ ├── roi_keypoint_predictors.py │ │ │ ├── roi_keypoint_feature_extractors.py │ │ │ ├── keypoint_head.py │ │ │ └── inference.py │ │ ├── mask_head │ │ │ ├── __init__.py │ │ │ ├── __pycache__ │ │ │ │ ├── loss.cpython-36.pyc │ │ │ │ ├── __init__.cpython-36.pyc │ │ │ │ ├── inference.cpython-36.pyc │ │ │ │ ├── mask_head.cpython-36.pyc │ │ │ │ ├── roi_mask_predictors.cpython-36.pyc │ │ │ │ └── roi_mask_feature_extractors.cpython-36.pyc │ │ │ ├── roi_mask_predictors.py │ │ │ ├── roi_mask_feature_extractors.py │ │ │ └── mask_head.py │ │ ├── __pycache__ │ │ │ ├── __init__.cpython-36.pyc │ │ │ └── roi_heads.cpython-36.pyc │ │ └── roi_heads.py │ ├── rpn │ │ ├── retinanet │ │ │ ├── __init__.py │ │ │ ├── __pycache__ │ │ │ │ ├── loss.cpython-36.pyc │ │ │ │ ├── __init__.cpython-36.pyc │ │ │ │ ├── inference.cpython-36.pyc │ │ │ │ └── retinanet.cpython-36.pyc │ │ │ └── loss.py │ │ ├── __init__.py │ │ ├── __pycache__ │ │ │ ├── loss.cpython-36.pyc │ │ │ ├── rpn.cpython-36.pyc │ │ │ ├── utils.cpython-36.pyc │ │ │ ├── __init__.cpython-36.pyc │ │ │ ├── inference.cpython-36.pyc │ │ │ └── anchor_generator.cpython-36.pyc │ │ └── utils.py │ ├── __pycache__ │ │ ├── utils.cpython-36.pyc │ │ ├── __init__.cpython-36.pyc │ │ ├── matcher.cpython-36.pyc │ │ ├── poolers.cpython-36.pyc │ │ ├── registry.cpython-36.pyc │ │ ├── box_coder.cpython-36.pyc │ │ ├── make_layers.cpython-36.pyc │ │ └── balanced_positive_negative_sampler.cpython-36.pyc │ ├── detector │ │ ├── __init__.py │ │ ├── __pycache__ │ │ │ ├── __init__.cpython-36.pyc │ │ │ ├── detectors.cpython-36.pyc │ │ │ └── generalized_rcnn.cpython-36.pyc │ │ ├── detectors.py │ │ └── generalized_rcnn.py │ ├── backbone │ │ ├── __pycache__ │ │ │ ├── fpn.cpython-36.pyc │ │ │ ├── fbnet.cpython-36.pyc │ │ │ ├── resnet.cpython-36.pyc │ │ │ ├── __init__.cpython-36.pyc │ │ │ ├── backbone.cpython-36.pyc │ │ │ ├── fbnet_builder.cpython-36.pyc │ │ │ └── fbnet_modeldef.cpython-36.pyc │ │ ├── __init__.py │ │ ├── backbone.py │ │ └── fpn.py │ ├── registry.py │ ├── utils.py │ ├── balanced_positive_negative_sampler.py │ ├── box_coder.py │ ├── make_layers.py │ └── poolers.py ├── structures │ ├── __init__.py │ ├── __pycache__ │ │ ├── __init__.cpython-36.pyc │ │ ├── keypoint.cpython-36.pyc │ │ ├── boxlist_ops.cpython-36.pyc │ │ ├── image_list.cpython-36.pyc │ │ ├── bounding_box.cpython-36.pyc │ │ └── segmentation_mask.cpython-36.pyc │ ├── image_list.py │ └── boxlist_ops.py ├── __init__.py ├── engine │ ├── __init__.py │ ├── inference.py │ ├── trainer.py │ └── bbox_aug.py ├── config │ ├── __init__.py │ └── __pycache__ │ │ ├── __init__.cpython-36.pyc │ │ └── defaults.cpython-36.pyc ├── data │ ├── __init__.py │ ├── __pycache__ │ │ ├── build.cpython-36.pyc │ │ ├── __init__.cpython-36.pyc │ │ └── collate_batch.cpython-36.pyc │ ├── datasets │ │ ├── __pycache__ │ │ │ ├── voc.cpython-36.pyc │ │ │ ├── coco.cpython-36.pyc │ │ │ ├── __init__.cpython-36.pyc │ │ │ └── concat_dataset.cpython-36.pyc │ │ ├── __init__.py │ │ ├── evaluation │ │ │ ├── voc │ │ │ │ └── __init__.py │ │ │ ├── coco │ │ │ │ └── __init__.py │ │ │ └── __init__.py │ │ ├── concat_dataset.py │ │ ├── list_dataset.py │ │ ├── coco.py │ │ └── voc.py │ ├── samplers │ │ ├── __pycache__ │ │ │ ├── __init__.cpython-36.pyc │ │ │ ├── distributed.cpython-36.pyc │ │ │ ├── grouped_batch_sampler.cpython-36.pyc │ │ │ └── iteration_based_batch_sampler.cpython-36.pyc │ │ ├── __init__.py │ │ ├── iteration_based_batch_sampler.py │ │ ├── distributed.py │ │ └── grouped_batch_sampler.py │ ├── transforms │ │ ├── __pycache__ │ │ │ ├── build.cpython-36.pyc │ │ │ ├── __init__.cpython-36.pyc │ │ │ └── transforms.cpython-36.pyc │ │ ├── __init__.py │ │ ├── build.py │ │ └── transforms.py │ ├── collate_batch.py │ └── README.md ├── layers │ ├── dcn │ │ ├── __init__.py │ │ ├── __pycache__ │ │ │ ├── __init__.cpython-36.pyc │ │ │ ├── deform_conv_func.cpython-36.pyc │ │ │ ├── deform_pool_func.cpython-36.pyc │ │ │ ├── deform_conv_module.cpython-36.pyc │ │ │ └── deform_pool_module.cpython-36.pyc │ │ └── deform_pool_func.py │ ├── __pycache__ │ │ ├── misc.cpython-36.pyc │ │ ├── nms.cpython-36.pyc │ │ ├── __init__.cpython-36.pyc │ │ ├── batch_norm.cpython-36.pyc │ │ ├── roi_align.cpython-36.pyc │ │ ├── roi_pool.cpython-36.pyc │ │ ├── smooth_l1_loss.cpython-36.pyc │ │ └── sigmoid_focal_loss.cpython-36.pyc │ ├── nms.py │ ├── smooth_l1_loss.py │ ├── batch_norm.py │ ├── _utils.py │ ├── __init__.py │ ├── roi_pool.py │ ├── roi_align.py │ └── sigmoid_focal_loss.py ├── solver │ ├── __init__.py │ ├── build.py │ └── lr_scheduler.py ├── csrc │ ├── cpu │ │ ├── vision.h │ │ └── nms_cpu.cpp │ ├── nms.h │ ├── SigmoidFocalLoss.h │ ├── vision.cpp │ ├── ROIPool.h │ ├── ROIAlign.h │ ├── deform_pool.h │ ├── cuda │ │ └── deform_pool_cuda.cu │ └── deform_conv.h └── setup.py ├── configs └── rcnn │ └── pretrained_vg_cfg.yaml ├── .gitignore └── README.md /maskrcnn_benchmark/utils/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/structures/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/retinanet/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/box_head/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/engine/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/config/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from .defaults import _C as cfg 3 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from .build import make_data_loader 3 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | # from .rpn import build_rpn 3 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/dcn/__init__.py: -------------------------------------------------------------------------------- 1 | # 2 | # Copied From [mmdetection](https://github.com/open-mmlab/mmdetection/tree/master/mmdet/ops/dcn) 3 | # 4 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/__pycache__/build.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/__pycache__/build.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/__pycache__/misc.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/__pycache__/misc.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/__pycache__/nms.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/__pycache__/nms.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/comm.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/comm.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/__pycache__/utils.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/__pycache__/utils.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/detector/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from .detectors import build_detection_model 3 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/cv2_util.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/cv2_util.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/imports.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/imports.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/registry.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/registry.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/config/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/config/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/config/__pycache__/defaults.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/config/__pycache__/defaults.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/__pycache__/voc.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/datasets/__pycache__/voc.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/__pycache__/batch_norm.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/__pycache__/batch_norm.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/__pycache__/roi_align.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/__pycache__/roi_align.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/__pycache__/roi_pool.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/__pycache__/roi_pool.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/__pycache__/matcher.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/__pycache__/matcher.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/__pycache__/poolers.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/__pycache__/poolers.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/__pycache__/registry.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/__pycache__/registry.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/__pycache__/loss.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/__pycache__/loss.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/__pycache__/rpn.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/__pycache__/rpn.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/checkpoint.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/checkpoint.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/model_zoo.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/model_zoo.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/__pycache__/collate_batch.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/__pycache__/collate_batch.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/__pycache__/coco.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/datasets/__pycache__/coco.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/dcn/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/dcn/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/__pycache__/box_coder.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/__pycache__/box_coder.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/__pycache__/utils.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/__pycache__/utils.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/structures/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/structures/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/structures/__pycache__/keypoint.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/structures/__pycache__/keypoint.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/miscellaneous.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/miscellaneous.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/datasets/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/samplers/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/samplers/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/transforms/__pycache__/build.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/transforms/__pycache__/build.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/__pycache__/smooth_l1_loss.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/__pycache__/smooth_l1_loss.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/__pycache__/make_layers.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/__pycache__/make_layers.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/__pycache__/fpn.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/backbone/__pycache__/fpn.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/__pycache__/inference.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/__pycache__/inference.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/structures/__pycache__/boxlist_ops.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/structures/__pycache__/boxlist_ops.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/structures/__pycache__/image_list.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/structures/__pycache__/image_list.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/c2_model_loading.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/c2_model_loading.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/transforms/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/transforms/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from .backbone import build_backbone 3 | from . import fbnet 4 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/__pycache__/fbnet.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/backbone/__pycache__/fbnet.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/__pycache__/resnet.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/backbone/__pycache__/resnet.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/structures/__pycache__/bounding_box.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/structures/__pycache__/bounding_box.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/samplers/__pycache__/distributed.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/samplers/__pycache__/distributed.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/transforms/__pycache__/transforms.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/transforms/__pycache__/transforms.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/__pycache__/sigmoid_focal_loss.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/__pycache__/sigmoid_focal_loss.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/dcn/__pycache__/deform_conv_func.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/dcn/__pycache__/deform_conv_func.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/dcn/__pycache__/deform_pool_func.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/dcn/__pycache__/deform_pool_func.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/backbone/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/__pycache__/backbone.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/backbone/__pycache__/backbone.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/detector/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/detector/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/detector/__pycache__/detectors.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/detector/__pycache__/detectors.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/retinanet/__pycache__/loss.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/retinanet/__pycache__/loss.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/__pycache__/model_serialization.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/utils/__pycache__/model_serialization.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/__pycache__/concat_dataset.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/datasets/__pycache__/concat_dataset.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/dcn/__pycache__/deform_conv_module.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/dcn/__pycache__/deform_conv_module.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/dcn/__pycache__/deform_pool_module.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/layers/dcn/__pycache__/deform_pool_module.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/__pycache__/roi_heads.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/__pycache__/roi_heads.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/__pycache__/anchor_generator.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/__pycache__/anchor_generator.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/structures/__pycache__/segmentation_mask.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/structures/__pycache__/segmentation_mask.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/__pycache__/fbnet_builder.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/backbone/__pycache__/fbnet_builder.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/__pycache__/fbnet_modeldef.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/backbone/__pycache__/fbnet_modeldef.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/loss.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/loss.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/retinanet/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/retinanet/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/retinanet/__pycache__/inference.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/retinanet/__pycache__/inference.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/retinanet/__pycache__/retinanet.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/rpn/retinanet/__pycache__/retinanet.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/detector/__pycache__/generalized_rcnn.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/detector/__pycache__/generalized_rcnn.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/loss.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/loss.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/samplers/__pycache__/grouped_batch_sampler.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/samplers/__pycache__/grouped_batch_sampler.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/box_head.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/box_head.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/inference.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/inference.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/loss.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/loss.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/inference.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/inference.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/mask_head.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/mask_head.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/__init__.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/__init__.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/inference.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/inference.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/samplers/__pycache__/iteration_based_batch_sampler.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/data/samplers/__pycache__/iteration_based_batch_sampler.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/__pycache__/balanced_positive_negative_sampler.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/__pycache__/balanced_positive_negative_sampler.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/README.md: -------------------------------------------------------------------------------- 1 | # Utility functions 2 | 3 | This folder contain utility functions that are not used in the 4 | core library, but are useful for building models or training 5 | code using the config system. 6 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/roi_box_predictors.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/roi_box_predictors.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/keypoint_head.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/keypoint_head.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/roi_mask_predictors.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/roi_mask_predictors.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/solver/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from .build import make_optimizer 3 | from .build import make_lr_scheduler 4 | from .lr_scheduler import WarmupMultiStepLR 5 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/roi_box_feature_extractors.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/box_head/__pycache__/roi_box_feature_extractors.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/roi_keypoint_predictors.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/roi_keypoint_predictors.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/roi_mask_feature_extractors.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/mask_head/__pycache__/roi_mask_feature_extractors.cpython-36.pyc -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/roi_keypoint_feature_extractors.cpython-36.pyc: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/ceyzaguirre4/NSM/HEAD/maskrcnn_benchmark/modeling/roi_heads/keypoint_head/__pycache__/roi_keypoint_feature_extractors.cpython-36.pyc -------------------------------------------------------------------------------- /configs/rcnn/pretrained_vg_cfg.yaml: -------------------------------------------------------------------------------- 1 | MODEL: 2 | WEIGHT: "weights/rcnn/faster_rcnn_ckpt.pth" 3 | BACKBONE: 4 | CONV_BODY: "R-101-C4" 5 | ROI_BOX_HEAD: 6 | NUM_CLASSES: 151 7 | RPN: 8 | MIN_SIZE: 4 9 | INPUT: 10 | MAX_SIZE_TRAIN: 1024 11 | MAX_SIZE_TEST: 1024 12 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from .coco import COCODataset 3 | from .voc import PascalVOCDataset 4 | from .concat_dataset import ConcatDataset 5 | 6 | __all__ = ["COCODataset", "ConcatDataset", "PascalVOCDataset"] 7 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/transforms/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from .transforms import Compose 3 | from .transforms import Resize 4 | from .transforms import RandomHorizontalFlip 5 | from .transforms import ToTensor 6 | from .transforms import Normalize 7 | 8 | from .build import build_transforms 9 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/samplers/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from .distributed import DistributedSampler 3 | from .grouped_batch_sampler import GroupedBatchSampler 4 | from .iteration_based_batch_sampler import IterationBasedBatchSampler 5 | 6 | __all__ = ["DistributedSampler", "GroupedBatchSampler", "IterationBasedBatchSampler"] 7 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/nms.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | # from ._utils import _C 3 | from maskrcnn_benchmark import _C 4 | 5 | from apex import amp 6 | 7 | # Only valid with fp32 inputs - give AMP the hint 8 | nms = amp.float_function(_C.nms) 9 | 10 | # nms.__doc__ = """ 11 | # This function performs Non-maximum suppresion""" 12 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/detector/detectors.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from .generalized_rcnn import GeneralizedRCNN 3 | 4 | 5 | _DETECTION_META_ARCHITECTURES = {"GeneralizedRCNN": GeneralizedRCNN} 6 | 7 | 8 | def build_detection_model(cfg): 9 | meta_arch = _DETECTION_META_ARCHITECTURES[cfg.MODEL.META_ARCHITECTURE] 10 | return meta_arch(cfg) 11 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/collect_env.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import PIL 3 | 4 | from torch.utils.collect_env import get_pretty_env_info 5 | 6 | 7 | def get_pil_version(): 8 | return "\n Pillow ({})".format(PIL.__version__) 9 | 10 | 11 | def collect_env_info(): 12 | env_str = get_pretty_env_info() 13 | env_str += get_pil_version() 14 | return env_str 15 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/registry.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | 3 | from maskrcnn_benchmark.utils.registry import Registry 4 | 5 | BACKBONES = Registry() 6 | RPN_HEADS = Registry() 7 | ROI_BOX_FEATURE_EXTRACTORS = Registry() 8 | ROI_BOX_PREDICTOR = Registry() 9 | ROI_KEYPOINT_FEATURE_EXTRACTORS = Registry() 10 | ROI_KEYPOINT_PREDICTOR = Registry() 11 | ROI_MASK_FEATURE_EXTRACTORS = Registry() 12 | ROI_MASK_PREDICTOR = Registry() 13 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/utils.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | """ 3 | Miscellaneous utility functions 4 | """ 5 | 6 | import torch 7 | 8 | 9 | def cat(tensors, dim=0): 10 | """ 11 | Efficient version of torch.cat that avoids a copy if there is only a single element in a list 12 | """ 13 | assert isinstance(tensors, (list, tuple)) 14 | if len(tensors) == 1: 15 | return tensors[0] 16 | return torch.cat(tensors, dim) 17 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | # compilation and distribution 2 | __pycache__ 3 | _ext 4 | *.pyc 5 | *.so 6 | *.egg-info/ 7 | dist/ 8 | 9 | # pytorch/python/numpy formats 10 | *.pth 11 | *.pkl 12 | *.npy 13 | 14 | # ipython/jupyter notebooks 15 | *.ipynb 16 | mrcnn-test.ipynb 17 | **/.ipynb_checkpoints/ 18 | 19 | # Editor temporaries 20 | *.swn 21 | *.swo 22 | *.swp 23 | *~ 24 | 25 | # vscode editor settings 26 | .vscode 27 | 28 | # MacOS 29 | .DS_Store 30 | 31 | # project dirs 32 | maskrcnn_benchmark/build 33 | _C.cpython-36m-x86_64-linux-gnu.so 34 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/smooth_l1_loss.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | 4 | 5 | # TODO maybe push this to nn? 6 | def smooth_l1_loss(input, target, beta=1. / 9, size_average=True): 7 | """ 8 | very similar to the smooth_l1_loss from pytorch, but with 9 | the extra beta parameter 10 | """ 11 | n = torch.abs(input - target) 12 | cond = n < beta 13 | loss = torch.where(cond, 0.5 * n ** 2 / beta, n - 0.5 * beta) 14 | if size_average: 15 | return loss.mean() 16 | return loss.sum() 17 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/evaluation/voc/__init__.py: -------------------------------------------------------------------------------- 1 | import logging 2 | 3 | from .voc_eval import do_voc_evaluation 4 | 5 | 6 | def voc_evaluation(dataset, predictions, output_folder, box_only, **_): 7 | logger = logging.getLogger("maskrcnn_benchmark.inference") 8 | if box_only: 9 | logger.warning("voc evaluation doesn't support box_only, ignored.") 10 | logger.info("performing voc evaluation, ignored iou_types.") 11 | return do_voc_evaluation( 12 | dataset=dataset, 13 | predictions=predictions, 14 | output_folder=output_folder, 15 | logger=logger, 16 | ) 17 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/evaluation/coco/__init__.py: -------------------------------------------------------------------------------- 1 | from .coco_eval import do_coco_evaluation 2 | 3 | 4 | def coco_evaluation( 5 | dataset, 6 | predictions, 7 | output_folder, 8 | box_only, 9 | iou_types, 10 | expected_results, 11 | expected_results_sigma_tol, 12 | ): 13 | return do_coco_evaluation( 14 | dataset=dataset, 15 | predictions=predictions, 16 | box_only=box_only, 17 | output_folder=output_folder, 18 | iou_types=iou_types, 19 | expected_results=expected_results, 20 | expected_results_sigma_tol=expected_results_sigma_tol, 21 | ) 22 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/cpu/vision.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #pragma once 3 | #include 4 | 5 | 6 | at::Tensor ROIAlign_forward_cpu(const at::Tensor& input, 7 | const at::Tensor& rois, 8 | const float spatial_scale, 9 | const int pooled_height, 10 | const int pooled_width, 11 | const int sampling_ratio); 12 | 13 | 14 | at::Tensor nms_cpu(const at::Tensor& dets, 15 | const at::Tensor& scores, 16 | const float threshold); 17 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/cv2_util.py: -------------------------------------------------------------------------------- 1 | """ 2 | Module for cv2 utility functions and maintaining version compatibility 3 | between 3.x and 4.x 4 | """ 5 | import cv2 6 | 7 | 8 | def findContours(*args, **kwargs): 9 | """ 10 | Wraps cv2.findContours to maintain compatiblity between versions 11 | 3 and 4 12 | 13 | Returns: 14 | contours, hierarchy 15 | """ 16 | if cv2.__version__.startswith('4'): 17 | contours, hierarchy = cv2.findContours(*args, **kwargs) 18 | elif cv2.__version__.startswith('3'): 19 | _, contours, hierarchy = cv2.findContours(*args, **kwargs) 20 | else: 21 | raise AssertionError( 22 | 'cv2 must be either version 3 or 4 to call this method') 23 | 24 | return contours, hierarchy 25 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/nms.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #pragma once 3 | #include "cpu/vision.h" 4 | 5 | #ifdef WITH_CUDA 6 | #include "cuda/vision.h" 7 | #endif 8 | 9 | 10 | at::Tensor nms(const at::Tensor& dets, 11 | const at::Tensor& scores, 12 | const float threshold) { 13 | 14 | if (dets.type().is_cuda()) { 15 | #ifdef WITH_CUDA 16 | // TODO raise error if not compiled with CUDA 17 | if (dets.numel() == 0) 18 | return at::empty({0}, dets.options().dtype(at::kLong).device(at::kCPU)); 19 | auto b = at::cat({dets, scores.unsqueeze(1)}, 1); 20 | return nms_cuda(b, threshold); 21 | #else 22 | AT_ERROR("Not compiled with GPU support"); 23 | #endif 24 | } 25 | 26 | at::Tensor result = nms_cpu(dets, scores, threshold); 27 | return result; 28 | } 29 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/concat_dataset.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import bisect 3 | 4 | from torch.utils.data.dataset import ConcatDataset as _ConcatDataset 5 | 6 | 7 | class ConcatDataset(_ConcatDataset): 8 | """ 9 | Same as torch.utils.data.dataset.ConcatDataset, but exposes an extra 10 | method for querying the sizes of the image 11 | """ 12 | 13 | def get_idxs(self, idx): 14 | dataset_idx = bisect.bisect_right(self.cumulative_sizes, idx) 15 | if dataset_idx == 0: 16 | sample_idx = idx 17 | else: 18 | sample_idx = idx - self.cumulative_sizes[dataset_idx - 1] 19 | return dataset_idx, sample_idx 20 | 21 | def get_img_info(self, idx): 22 | dataset_idx, sample_idx = self.get_idxs(idx) 23 | return self.datasets[dataset_idx].get_img_info(sample_idx) 24 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/logger.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import logging 3 | import os 4 | import sys 5 | 6 | 7 | def setup_logger(name, save_dir, distributed_rank, filename="log.txt"): 8 | logger = logging.getLogger(name) 9 | logger.setLevel(logging.DEBUG) 10 | # don't log results for the non-master process 11 | if distributed_rank > 0: 12 | return logger 13 | ch = logging.StreamHandler(stream=sys.stdout) 14 | ch.setLevel(logging.DEBUG) 15 | formatter = logging.Formatter("%(asctime)s %(name)s %(levelname)s: %(message)s") 16 | ch.setFormatter(formatter) 17 | logger.addHandler(ch) 18 | 19 | if save_dir: 20 | fh = logging.FileHandler(os.path.join(save_dir, filename)) 21 | fh.setLevel(logging.DEBUG) 22 | fh.setFormatter(formatter) 23 | logger.addHandler(fh) 24 | 25 | return logger 26 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/imports.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | 4 | if torch._six.PY3: 5 | import importlib 6 | import importlib.util 7 | import sys 8 | 9 | 10 | # from https://stackoverflow.com/questions/67631/how-to-import-a-module-given-the-full-path?utm_medium=organic&utm_source=google_rich_qa&utm_campaign=google_rich_qa 11 | def import_file(module_name, file_path, make_importable=False): 12 | spec = importlib.util.spec_from_file_location(module_name, file_path) 13 | module = importlib.util.module_from_spec(spec) 14 | spec.loader.exec_module(module) 15 | if make_importable: 16 | sys.modules[module_name] = module 17 | return module 18 | else: 19 | import imp 20 | 21 | def import_file(module_name, file_path, make_importable=None): 22 | module = imp.load_source(module_name, file_path) 23 | return module 24 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/collate_batch.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from maskrcnn_benchmark.structures.image_list import to_image_list 3 | 4 | 5 | class BatchCollator(object): 6 | """ 7 | From a list of samples from the dataset, 8 | returns the batched images and targets. 9 | This should be passed to the DataLoader 10 | """ 11 | 12 | def __init__(self, size_divisible=0): 13 | self.size_divisible = size_divisible 14 | 15 | def __call__(self, batch): 16 | transposed_batch = list(zip(*batch)) 17 | images = to_image_list(transposed_batch[0], self.size_divisible) 18 | targets = transposed_batch[1] 19 | img_ids = transposed_batch[2] 20 | return images, targets, img_ids 21 | 22 | 23 | class BBoxAugCollator(object): 24 | """ 25 | From a list of samples from the dataset, 26 | returns the images and targets. 27 | Images should be converted to batched images in `im_detect_bbox_aug` 28 | """ 29 | 30 | def __call__(self, batch): 31 | return list(zip(*batch)) 32 | 33 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/solver/build.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | 4 | from .lr_scheduler import WarmupMultiStepLR 5 | 6 | 7 | def make_optimizer(cfg, model): 8 | params = [] 9 | for key, value in model.named_parameters(): 10 | if not value.requires_grad: 11 | continue 12 | lr = cfg.SOLVER.BASE_LR 13 | weight_decay = cfg.SOLVER.WEIGHT_DECAY 14 | if "bias" in key: 15 | lr = cfg.SOLVER.BASE_LR * cfg.SOLVER.BIAS_LR_FACTOR 16 | weight_decay = cfg.SOLVER.WEIGHT_DECAY_BIAS 17 | params += [{"params": [value], "lr": lr, "weight_decay": weight_decay}] 18 | 19 | optimizer = torch.optim.SGD(params, lr, momentum=cfg.SOLVER.MOMENTUM) 20 | return optimizer 21 | 22 | 23 | def make_lr_scheduler(cfg, optimizer): 24 | return WarmupMultiStepLR( 25 | optimizer, 26 | cfg.SOLVER.STEPS, 27 | cfg.SOLVER.GAMMA, 28 | warmup_factor=cfg.SOLVER.WARMUP_FACTOR, 29 | warmup_iters=cfg.SOLVER.WARMUP_ITERS, 30 | warmup_method=cfg.SOLVER.WARMUP_METHOD, 31 | ) 32 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/list_dataset.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | """ 3 | Simple dataset class that wraps a list of path names 4 | """ 5 | 6 | from PIL import Image 7 | 8 | from maskrcnn_benchmark.structures.bounding_box import BoxList 9 | 10 | 11 | class ListDataset(object): 12 | def __init__(self, image_lists, transforms=None): 13 | self.image_lists = image_lists 14 | self.transforms = transforms 15 | 16 | def __getitem__(self, item): 17 | img = Image.open(self.image_lists[item]).convert("RGB") 18 | 19 | # dummy target 20 | w, h = img.size 21 | target = BoxList([[0, 0, w, h]], img.size, mode="xyxy") 22 | 23 | if self.transforms is not None: 24 | img, target = self.transforms(img, target) 25 | 26 | return img, target 27 | 28 | def __len__(self): 29 | return len(self.image_lists) 30 | 31 | def get_img_info(self, item): 32 | """ 33 | Return the image dimensions for the image, without 34 | loading and pre-processing it 35 | """ 36 | pass 37 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/evaluation/__init__.py: -------------------------------------------------------------------------------- 1 | from maskrcnn_benchmark.data import datasets 2 | 3 | from .coco import coco_evaluation 4 | from .voc import voc_evaluation 5 | 6 | 7 | def evaluate(dataset, predictions, output_folder, **kwargs): 8 | """evaluate dataset using different methods based on dataset type. 9 | Args: 10 | dataset: Dataset object 11 | predictions(list[BoxList]): each item in the list represents the 12 | prediction results for one image. 13 | output_folder: output folder, to save evaluation files or results. 14 | **kwargs: other args. 15 | Returns: 16 | evaluation result 17 | """ 18 | args = dict( 19 | dataset=dataset, predictions=predictions, output_folder=output_folder, **kwargs 20 | ) 21 | if isinstance(dataset, datasets.COCODataset): 22 | return coco_evaluation(**args) 23 | elif isinstance(dataset, datasets.PascalVOCDataset): 24 | return voc_evaluation(**args) 25 | else: 26 | dataset_name = dataset.__class__.__name__ 27 | raise NotImplementedError("Unsupported dataset type {}.".format(dataset_name)) 28 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # NSM 2 | 3 | Neural State Machine implemented in [PyTorch](http://pytorch.org/) as presented [here](https://arxiv.org/abs/1907.03950). 4 | This is the first code implementation of the model and is based on V1 of the paper on [arxiv](https://arxiv.org). 5 | As can be expected, the code is incomplete and makes several assumptions where the paper wasn't clear enough. 6 | For the time being this code is not ready to run and several steps are needed for it to train on any VQA dataset. 7 | Principal among these is the need for a functioning [graph-rcnn](https://arxiv.org/pdf/1808.00191.pdf) to generate the scene graphs. 8 | 9 | In the meantime, I hope the code helps readers understand better the paper, and I'm open to any collaborators who wish to help with features or efficiency. 10 | 11 | ## RCNN compilation 12 | 13 | We use Facebook's implementation of maskrcnn: [maskrcnn-benchmark](https://github.com/facebookresearch/maskrcnn-benchmark). To compile this code use the following snippet. 14 | 15 | ``` 16 | # pytorch, apex and maskrcnn_benchmark must be compiled with the same version of CUDA. 17 | cd maskrcnn_benchmark 18 | python setup.py build develop 19 | ``` 20 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/SigmoidFocalLoss.h: -------------------------------------------------------------------------------- 1 | #pragma once 2 | 3 | #include "cpu/vision.h" 4 | 5 | #ifdef WITH_CUDA 6 | #include "cuda/vision.h" 7 | #endif 8 | 9 | // Interface for Python 10 | at::Tensor SigmoidFocalLoss_forward( 11 | const at::Tensor& logits, 12 | const at::Tensor& targets, 13 | const int num_classes, 14 | const float gamma, 15 | const float alpha) { 16 | if (logits.type().is_cuda()) { 17 | #ifdef WITH_CUDA 18 | return SigmoidFocalLoss_forward_cuda(logits, targets, num_classes, gamma, alpha); 19 | #else 20 | AT_ERROR("Not compiled with GPU support"); 21 | #endif 22 | } 23 | AT_ERROR("Not implemented on the CPU"); 24 | } 25 | 26 | at::Tensor SigmoidFocalLoss_backward( 27 | const at::Tensor& logits, 28 | const at::Tensor& targets, 29 | const at::Tensor& d_losses, 30 | const int num_classes, 31 | const float gamma, 32 | const float alpha) { 33 | if (logits.type().is_cuda()) { 34 | #ifdef WITH_CUDA 35 | return SigmoidFocalLoss_backward_cuda(logits, targets, d_losses, num_classes, gamma, alpha); 36 | #else 37 | AT_ERROR("Not compiled with GPU support"); 38 | #endif 39 | } 40 | AT_ERROR("Not implemented on the CPU"); 41 | } 42 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/batch_norm.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | from torch import nn 4 | 5 | 6 | class FrozenBatchNorm2d(nn.Module): 7 | """ 8 | BatchNorm2d where the batch statistics and the affine parameters 9 | are fixed 10 | """ 11 | 12 | def __init__(self, n): 13 | super(FrozenBatchNorm2d, self).__init__() 14 | self.register_buffer("weight", torch.ones(n)) 15 | self.register_buffer("bias", torch.zeros(n)) 16 | self.register_buffer("running_mean", torch.zeros(n)) 17 | self.register_buffer("running_var", torch.ones(n)) 18 | 19 | def forward(self, x): 20 | # Cast all fixed parameters to half() if necessary 21 | if x.dtype == torch.float16: 22 | self.weight = self.weight.half() 23 | self.bias = self.bias.half() 24 | self.running_mean = self.running_mean.half() 25 | self.running_var = self.running_var.half() 26 | 27 | scale = self.weight * self.running_var.rsqrt() 28 | bias = self.bias - self.running_mean * scale 29 | scale = scale.reshape(1, -1, 1, 1) 30 | bias = bias.reshape(1, -1, 1, 1) 31 | return x * scale + bias 32 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/samplers/iteration_based_batch_sampler.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from torch.utils.data.sampler import BatchSampler 3 | 4 | 5 | class IterationBasedBatchSampler(BatchSampler): 6 | """ 7 | Wraps a BatchSampler, resampling from it until 8 | a specified number of iterations have been sampled 9 | """ 10 | 11 | def __init__(self, batch_sampler, num_iterations, start_iter=0): 12 | self.batch_sampler = batch_sampler 13 | self.num_iterations = num_iterations 14 | self.start_iter = start_iter 15 | 16 | def __iter__(self): 17 | iteration = self.start_iter 18 | while iteration <= self.num_iterations: 19 | # if the underlying sampler has a set_epoch method, like 20 | # DistributedSampler, used for making each process see 21 | # a different split of the dataset, then set it 22 | if hasattr(self.batch_sampler.sampler, "set_epoch"): 23 | self.batch_sampler.sampler.set_epoch(iteration) 24 | for batch in self.batch_sampler: 25 | iteration += 1 26 | if iteration > self.num_iterations: 27 | break 28 | yield batch 29 | 30 | def __len__(self): 31 | return self.num_iterations 32 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/miscellaneous.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import errno 3 | import json 4 | import logging 5 | import os 6 | from .comm import is_main_process 7 | 8 | 9 | def mkdir(path): 10 | try: 11 | os.makedirs(path) 12 | except OSError as e: 13 | if e.errno != errno.EEXIST: 14 | raise 15 | 16 | 17 | def save_labels(dataset_list, output_dir): 18 | if is_main_process(): 19 | logger = logging.getLogger(__name__) 20 | 21 | ids_to_labels = {} 22 | for dataset in dataset_list: 23 | if hasattr(dataset, 'categories'): 24 | ids_to_labels.update(dataset.categories) 25 | else: 26 | logger.warning("Dataset [{}] has no categories attribute, labels.json file won't be created".format( 27 | dataset.__class__.__name__)) 28 | 29 | if ids_to_labels: 30 | labels_file = os.path.join(output_dir, 'labels.json') 31 | logger.info("Saving labels mapping into {}".format(labels_file)) 32 | with open(labels_file, 'w') as f: 33 | json.dump(ids_to_labels, f, indent=2) 34 | 35 | 36 | def save_config(cfg, path): 37 | if is_main_process(): 38 | with open(path, 'w') as f: 39 | f.write(cfg.dump()) 40 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/timer.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | 3 | 4 | import time 5 | import datetime 6 | 7 | 8 | class Timer(object): 9 | def __init__(self): 10 | self.reset() 11 | 12 | @property 13 | def average_time(self): 14 | return self.total_time / self.calls if self.calls > 0 else 0.0 15 | 16 | def tic(self): 17 | # using time.time instead of time.clock because time time.clock 18 | # does not normalize for multithreading 19 | self.start_time = time.time() 20 | 21 | def toc(self, average=True): 22 | self.add(time.time() - self.start_time) 23 | if average: 24 | return self.average_time 25 | else: 26 | return self.diff 27 | 28 | def add(self, time_diff): 29 | self.diff = time_diff 30 | self.total_time += self.diff 31 | self.calls += 1 32 | 33 | def reset(self): 34 | self.total_time = 0.0 35 | self.calls = 0 36 | self.start_time = 0.0 37 | self.diff = 0.0 38 | 39 | def avg_time_str(self): 40 | time_str = str(datetime.timedelta(seconds=self.average_time)) 41 | return time_str 42 | 43 | 44 | def get_time_str(time_diff): 45 | time_str = str(datetime.timedelta(seconds=time_diff)) 46 | return time_str 47 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/_utils.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import glob 3 | import os.path 4 | 5 | import torch 6 | 7 | try: 8 | from torch.utils.cpp_extension import load as load_ext 9 | from torch.utils.cpp_extension import CUDA_HOME 10 | except ImportError: 11 | raise ImportError("The cpp layer extensions requires PyTorch 0.4 or higher") 12 | 13 | 14 | def _load_C_extensions(): 15 | this_dir = os.path.dirname(os.path.abspath(__file__)) 16 | this_dir = os.path.dirname(this_dir) 17 | this_dir = os.path.join(this_dir, "csrc") 18 | 19 | main_file = glob.glob(os.path.join(this_dir, "*.cpp")) 20 | source_cpu = glob.glob(os.path.join(this_dir, "cpu", "*.cpp")) 21 | source_cuda = glob.glob(os.path.join(this_dir, "cuda", "*.cu")) 22 | 23 | source = main_file + source_cpu 24 | 25 | extra_cflags = [] 26 | if torch.cuda.is_available() and CUDA_HOME is not None: 27 | source.extend(source_cuda) 28 | extra_cflags = ["-DWITH_CUDA"] 29 | source = [os.path.join(this_dir, s) for s in source] 30 | extra_include_paths = [this_dir] 31 | return load_ext( 32 | "torchvision", 33 | source, 34 | extra_cflags=extra_cflags, 35 | extra_include_paths=extra_include_paths, 36 | ) 37 | 38 | 39 | _C = _load_C_extensions() 40 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/env.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import os 3 | 4 | from maskrcnn_benchmark.utils.imports import import_file 5 | 6 | 7 | def setup_environment(): 8 | """Perform environment setup work. The default setup is a no-op, but this 9 | function allows the user to specify a Python source file that performs 10 | custom setup work that may be necessary to their computing environment. 11 | """ 12 | custom_module_path = os.environ.get("TORCH_DETECTRON_ENV_MODULE") 13 | if custom_module_path: 14 | setup_custom_environment(custom_module_path) 15 | else: 16 | # The default setup is a no-op 17 | pass 18 | 19 | 20 | def setup_custom_environment(custom_module_path): 21 | """Load custom environment setup from a Python source file and run the setup 22 | function. 23 | """ 24 | module = import_file("maskrcnn_benchmark.utils.env.custom_module", custom_module_path) 25 | assert hasattr(module, "setup_environment") and callable( 26 | module.setup_environment 27 | ), ( 28 | "Custom environment module defined in {} does not have the " 29 | "required callable attribute 'setup_environment'." 30 | ).format( 31 | custom_module_path 32 | ) 33 | module.setup_environment() 34 | 35 | 36 | # Force environment setup when this module is imported 37 | setup_environment() 38 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/roi_keypoint_predictors.py: -------------------------------------------------------------------------------- 1 | from torch import nn 2 | 3 | from maskrcnn_benchmark import layers 4 | from maskrcnn_benchmark.modeling import registry 5 | 6 | 7 | @registry.ROI_KEYPOINT_PREDICTOR.register("KeypointRCNNPredictor") 8 | class KeypointRCNNPredictor(nn.Module): 9 | def __init__(self, cfg, in_channels): 10 | super(KeypointRCNNPredictor, self).__init__() 11 | input_features = in_channels 12 | num_keypoints = cfg.MODEL.ROI_KEYPOINT_HEAD.NUM_CLASSES 13 | deconv_kernel = 4 14 | self.kps_score_lowres = layers.ConvTranspose2d( 15 | input_features, 16 | num_keypoints, 17 | deconv_kernel, 18 | stride=2, 19 | padding=deconv_kernel // 2 - 1, 20 | ) 21 | nn.init.kaiming_normal_( 22 | self.kps_score_lowres.weight, mode="fan_out", nonlinearity="relu" 23 | ) 24 | nn.init.constant_(self.kps_score_lowres.bias, 0) 25 | self.up_scale = 2 26 | self.out_channels = num_keypoints 27 | 28 | def forward(self, x): 29 | x = self.kps_score_lowres(x) 30 | x = layers.interpolate( 31 | x, scale_factor=self.up_scale, mode="bilinear", align_corners=False 32 | ) 33 | return x 34 | 35 | 36 | def make_roi_keypoint_predictor(cfg, in_channels): 37 | func = registry.ROI_KEYPOINT_PREDICTOR[cfg.MODEL.ROI_KEYPOINT_HEAD.PREDICTOR] 38 | return func(cfg, in_channels) 39 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/__init__.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | 4 | from .batch_norm import FrozenBatchNorm2d 5 | from .misc import Conv2d 6 | from .misc import DFConv2d 7 | from .misc import ConvTranspose2d 8 | from .misc import BatchNorm2d 9 | from .misc import interpolate 10 | from .nms import nms 11 | from .roi_align import ROIAlign 12 | from .roi_align import roi_align 13 | from .roi_pool import ROIPool 14 | from .roi_pool import roi_pool 15 | from .smooth_l1_loss import smooth_l1_loss 16 | from .sigmoid_focal_loss import SigmoidFocalLoss 17 | from .dcn.deform_conv_func import deform_conv, modulated_deform_conv 18 | from .dcn.deform_conv_module import DeformConv, ModulatedDeformConv, ModulatedDeformConvPack 19 | from .dcn.deform_pool_func import deform_roi_pooling 20 | from .dcn.deform_pool_module import DeformRoIPooling, DeformRoIPoolingPack, ModulatedDeformRoIPoolingPack 21 | 22 | 23 | __all__ = [ 24 | "nms", 25 | "roi_align", 26 | "ROIAlign", 27 | "roi_pool", 28 | "ROIPool", 29 | "smooth_l1_loss", 30 | "Conv2d", 31 | "DFConv2d", 32 | "ConvTranspose2d", 33 | "interpolate", 34 | "BatchNorm2d", 35 | "FrozenBatchNorm2d", 36 | "SigmoidFocalLoss", 37 | 'deform_conv', 38 | 'modulated_deform_conv', 39 | 'DeformConv', 40 | 'ModulatedDeformConv', 41 | 'ModulatedDeformConvPack', 42 | 'deform_roi_pooling', 43 | 'DeformRoIPooling', 44 | 'DeformRoIPoolingPack', 45 | 'ModulatedDeformRoIPoolingPack', 46 | ] 47 | 48 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/registry.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | 3 | 4 | def _register_generic(module_dict, module_name, module): 5 | assert module_name not in module_dict 6 | module_dict[module_name] = module 7 | 8 | 9 | class Registry(dict): 10 | ''' 11 | A helper class for managing registering modules, it extends a dictionary 12 | and provides a register functions. 13 | 14 | Eg. creeting a registry: 15 | some_registry = Registry({"default": default_module}) 16 | 17 | There're two ways of registering new modules: 18 | 1): normal way is just calling register function: 19 | def foo(): 20 | ... 21 | some_registry.register("foo_module", foo) 22 | 2): used as decorator when declaring the module: 23 | @some_registry.register("foo_module") 24 | @some_registry.register("foo_modeul_nickname") 25 | def foo(): 26 | ... 27 | 28 | Access of module is just like using a dictionary, eg: 29 | f = some_registry["foo_modeul"] 30 | ''' 31 | def __init__(self, *args, **kwargs): 32 | super(Registry, self).__init__(*args, **kwargs) 33 | 34 | def register(self, module_name, module=None): 35 | # used as function call 36 | if module is not None: 37 | _register_generic(self, module_name, module) 38 | return 39 | 40 | # used as decorator 41 | def register_fn(fn): 42 | _register_generic(self, module_name, fn) 43 | return fn 44 | 45 | return register_fn 46 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/vision.cpp: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #include "nms.h" 3 | #include "ROIAlign.h" 4 | #include "ROIPool.h" 5 | #include "SigmoidFocalLoss.h" 6 | #include "deform_conv.h" 7 | #include "deform_pool.h" 8 | 9 | PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { 10 | m.def("nms", &nms, "non-maximum suppression"); 11 | m.def("roi_align_forward", &ROIAlign_forward, "ROIAlign_forward"); 12 | m.def("roi_align_backward", &ROIAlign_backward, "ROIAlign_backward"); 13 | m.def("roi_pool_forward", &ROIPool_forward, "ROIPool_forward"); 14 | m.def("roi_pool_backward", &ROIPool_backward, "ROIPool_backward"); 15 | m.def("sigmoid_focalloss_forward", &SigmoidFocalLoss_forward, "SigmoidFocalLoss_forward"); 16 | m.def("sigmoid_focalloss_backward", &SigmoidFocalLoss_backward, "SigmoidFocalLoss_backward"); 17 | // dcn-v2 18 | m.def("deform_conv_forward", &deform_conv_forward, "deform_conv_forward"); 19 | m.def("deform_conv_backward_input", &deform_conv_backward_input, "deform_conv_backward_input"); 20 | m.def("deform_conv_backward_parameters", &deform_conv_backward_parameters, "deform_conv_backward_parameters"); 21 | m.def("modulated_deform_conv_forward", &modulated_deform_conv_forward, "modulated_deform_conv_forward"); 22 | m.def("modulated_deform_conv_backward", &modulated_deform_conv_backward, "modulated_deform_conv_backward"); 23 | m.def("deform_psroi_pooling_forward", &deform_psroi_pooling_forward, "deform_psroi_pooling_forward"); 24 | m.def("deform_psroi_pooling_backward", &deform_psroi_pooling_backward, "deform_psroi_pooling_backward"); 25 | } -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/transforms/build.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from . import transforms as T 3 | 4 | 5 | def build_transforms(cfg, is_train=True): 6 | if is_train: 7 | min_size = cfg.INPUT.MIN_SIZE_TRAIN 8 | max_size = cfg.INPUT.MAX_SIZE_TRAIN 9 | flip_horizontal_prob = 0.5 # cfg.INPUT.FLIP_PROB_TRAIN 10 | flip_vertical_prob = cfg.INPUT.VERTICAL_FLIP_PROB_TRAIN 11 | brightness = cfg.INPUT.BRIGHTNESS 12 | contrast = cfg.INPUT.CONTRAST 13 | saturation = cfg.INPUT.SATURATION 14 | hue = cfg.INPUT.HUE 15 | else: 16 | min_size = cfg.INPUT.MIN_SIZE_TEST 17 | max_size = cfg.INPUT.MAX_SIZE_TEST 18 | flip_horizontal_prob = 0.0 19 | flip_vertical_prob = 0.0 20 | brightness = 0.0 21 | contrast = 0.0 22 | saturation = 0.0 23 | hue = 0.0 24 | 25 | to_bgr255 = cfg.INPUT.TO_BGR255 26 | normalize_transform = T.Normalize( 27 | mean=cfg.INPUT.PIXEL_MEAN, std=cfg.INPUT.PIXEL_STD, to_bgr255=to_bgr255 28 | ) 29 | color_jitter = T.ColorJitter( 30 | brightness=brightness, 31 | contrast=contrast, 32 | saturation=saturation, 33 | hue=hue, 34 | ) 35 | 36 | transform = T.Compose( 37 | [ 38 | color_jitter, 39 | T.Resize(min_size, max_size), 40 | T.RandomHorizontalFlip(flip_horizontal_prob), 41 | T.RandomVerticalFlip(flip_vertical_prob), 42 | T.ToTensor(), 43 | normalize_transform, 44 | ] 45 | ) 46 | return transform 47 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/ROIPool.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 | 11 | std::tuple ROIPool_forward(const at::Tensor& input, 12 | const at::Tensor& rois, 13 | const float spatial_scale, 14 | const int pooled_height, 15 | const int pooled_width) { 16 | if (input.type().is_cuda()) { 17 | #ifdef WITH_CUDA 18 | return ROIPool_forward_cuda(input, rois, spatial_scale, pooled_height, pooled_width); 19 | #else 20 | AT_ERROR("Not compiled with GPU support"); 21 | #endif 22 | } 23 | AT_ERROR("Not implemented on the CPU"); 24 | } 25 | 26 | at::Tensor ROIPool_backward(const at::Tensor& grad, 27 | const at::Tensor& input, 28 | const at::Tensor& rois, 29 | const at::Tensor& argmax, 30 | const float spatial_scale, 31 | const int pooled_height, 32 | const int pooled_width, 33 | const int batch_size, 34 | const int channels, 35 | const int height, 36 | const int width) { 37 | if (grad.type().is_cuda()) { 38 | #ifdef WITH_CUDA 39 | return ROIPool_backward_cuda(grad, input, rois, argmax, spatial_scale, pooled_height, pooled_width, batch_size, channels, height, width); 40 | #else 41 | AT_ERROR("Not compiled with GPU support"); 42 | #endif 43 | } 44 | AT_ERROR("Not implemented on the CPU"); 45 | } 46 | 47 | 48 | 49 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/ROIAlign.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 ROIAlign_forward(const at::Tensor& input, 12 | const at::Tensor& rois, 13 | const float spatial_scale, 14 | const int pooled_height, 15 | const int pooled_width, 16 | const int sampling_ratio) { 17 | if (input.type().is_cuda()) { 18 | #ifdef WITH_CUDA 19 | return ROIAlign_forward_cuda(input, rois, spatial_scale, pooled_height, pooled_width, sampling_ratio); 20 | #else 21 | AT_ERROR("Not compiled with GPU support"); 22 | #endif 23 | } 24 | return ROIAlign_forward_cpu(input, rois, spatial_scale, pooled_height, pooled_width, sampling_ratio); 25 | } 26 | 27 | at::Tensor ROIAlign_backward(const at::Tensor& grad, 28 | const at::Tensor& rois, 29 | const float spatial_scale, 30 | const int pooled_height, 31 | const int pooled_width, 32 | const int batch_size, 33 | const int channels, 34 | const int height, 35 | const int width, 36 | const int sampling_ratio) { 37 | if (grad.type().is_cuda()) { 38 | #ifdef WITH_CUDA 39 | return ROIAlign_backward_cuda(grad, rois, spatial_scale, pooled_height, pooled_width, batch_size, channels, height, width, sampling_ratio); 40 | #else 41 | AT_ERROR("Not compiled with GPU support"); 42 | #endif 43 | } 44 | AT_ERROR("Not implemented on the CPU"); 45 | } 46 | 47 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/utils.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | """ 3 | Utility functions minipulating the prediction layers 4 | """ 5 | 6 | from ..utils import cat 7 | 8 | import torch 9 | 10 | def permute_and_flatten(layer, N, A, C, H, W): 11 | layer = layer.view(N, -1, C, H, W) 12 | layer = layer.permute(0, 3, 4, 1, 2) 13 | layer = layer.reshape(N, -1, C) 14 | return layer 15 | 16 | 17 | def concat_box_prediction_layers(box_cls, box_regression): 18 | box_cls_flattened = [] 19 | box_regression_flattened = [] 20 | # for each feature level, permute the outputs to make them be in the 21 | # same format as the labels. Note that the labels are computed for 22 | # all feature levels concatenated, so we keep the same representation 23 | # for the objectness and the box_regression 24 | for box_cls_per_level, box_regression_per_level in zip( 25 | box_cls, box_regression 26 | ): 27 | N, AxC, H, W = box_cls_per_level.shape 28 | Ax4 = box_regression_per_level.shape[1] 29 | A = Ax4 // 4 30 | C = AxC // A 31 | box_cls_per_level = permute_and_flatten( 32 | box_cls_per_level, N, A, C, H, W 33 | ) 34 | box_cls_flattened.append(box_cls_per_level) 35 | 36 | box_regression_per_level = permute_and_flatten( 37 | box_regression_per_level, N, A, 4, H, W 38 | ) 39 | box_regression_flattened.append(box_regression_per_level) 40 | # concatenate on the first dimension (representing the feature levels), to 41 | # take into account the way the labels were generated (with all feature maps 42 | # being concatenated as well) 43 | box_cls = cat(box_cls_flattened, dim=1).reshape(-1, C) 44 | box_regression = cat(box_regression_flattened, dim=1).reshape(-1, 4) 45 | return box_cls, box_regression 46 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/solver/lr_scheduler.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from bisect import bisect_right 3 | 4 | import torch 5 | 6 | 7 | # FIXME ideally this would be achieved with a CombinedLRScheduler, 8 | # separating MultiStepLR with WarmupLR 9 | # but the current LRScheduler design doesn't allow it 10 | class WarmupMultiStepLR(torch.optim.lr_scheduler._LRScheduler): 11 | def __init__( 12 | self, 13 | optimizer, 14 | milestones, 15 | gamma=0.1, 16 | warmup_factor=1.0 / 3, 17 | warmup_iters=500, 18 | warmup_method="linear", 19 | last_epoch=-1, 20 | ): 21 | if not list(milestones) == sorted(milestones): 22 | raise ValueError( 23 | "Milestones should be a list of" " increasing integers. Got {}", 24 | milestones, 25 | ) 26 | 27 | if warmup_method not in ("constant", "linear"): 28 | raise ValueError( 29 | "Only 'constant' or 'linear' warmup_method accepted" 30 | "got {}".format(warmup_method) 31 | ) 32 | self.milestones = milestones 33 | self.gamma = gamma 34 | self.warmup_factor = warmup_factor 35 | self.warmup_iters = warmup_iters 36 | self.warmup_method = warmup_method 37 | super(WarmupMultiStepLR, self).__init__(optimizer, last_epoch) 38 | 39 | def get_lr(self): 40 | warmup_factor = 1 41 | if self.last_epoch < self.warmup_iters: 42 | if self.warmup_method == "constant": 43 | warmup_factor = self.warmup_factor 44 | elif self.warmup_method == "linear": 45 | alpha = float(self.last_epoch) / self.warmup_iters 46 | warmup_factor = self.warmup_factor * (1 - alpha) + alpha 47 | return [ 48 | base_lr 49 | * warmup_factor 50 | * self.gamma ** bisect_right(self.milestones, self.last_epoch) 51 | for base_lr in self.base_lrs 52 | ] 53 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/deform_pool.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #pragma once 3 | #include "cpu/vision.h" 4 | 5 | #ifdef WITH_CUDA 6 | #include "cuda/vision.h" 7 | #endif 8 | 9 | 10 | // Interface for Python 11 | void deform_psroi_pooling_forward( 12 | at::Tensor input, 13 | at::Tensor bbox, 14 | at::Tensor trans, 15 | at::Tensor out, 16 | at::Tensor top_count, 17 | const int no_trans, 18 | const float spatial_scale, 19 | const int output_dim, 20 | const int group_size, 21 | const int pooled_size, 22 | const int part_size, 23 | const int sample_per_part, 24 | const float trans_std) 25 | { 26 | if (input.type().is_cuda()) { 27 | #ifdef WITH_CUDA 28 | return deform_psroi_pooling_cuda_forward( 29 | input, bbox, trans, out, top_count, 30 | no_trans, spatial_scale, output_dim, group_size, 31 | pooled_size, part_size, sample_per_part, trans_std 32 | ); 33 | #else 34 | AT_ERROR("Not compiled with GPU support"); 35 | #endif 36 | } 37 | AT_ERROR("Not implemented on the CPU"); 38 | } 39 | 40 | 41 | void deform_psroi_pooling_backward( 42 | at::Tensor out_grad, 43 | at::Tensor input, 44 | at::Tensor bbox, 45 | at::Tensor trans, 46 | at::Tensor top_count, 47 | at::Tensor input_grad, 48 | at::Tensor trans_grad, 49 | const int no_trans, 50 | const float spatial_scale, 51 | const int output_dim, 52 | const int group_size, 53 | const int pooled_size, 54 | const int part_size, 55 | const int sample_per_part, 56 | const float trans_std) 57 | { 58 | if (input.type().is_cuda()) { 59 | #ifdef WITH_CUDA 60 | return deform_psroi_pooling_cuda_backward( 61 | out_grad, input, bbox, trans, top_count, input_grad, trans_grad, 62 | no_trans, spatial_scale, output_dim, group_size, pooled_size, 63 | part_size, sample_per_part, trans_std 64 | ); 65 | #else 66 | AT_ERROR("Not compiled with GPU support"); 67 | #endif 68 | } 69 | AT_ERROR("Not implemented on the CPU"); 70 | } 71 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/roi_keypoint_feature_extractors.py: -------------------------------------------------------------------------------- 1 | from torch import nn 2 | from torch.nn import functional as F 3 | 4 | from maskrcnn_benchmark.modeling import registry 5 | from maskrcnn_benchmark.modeling.poolers import Pooler 6 | 7 | from maskrcnn_benchmark.layers import Conv2d 8 | 9 | 10 | @registry.ROI_KEYPOINT_FEATURE_EXTRACTORS.register("KeypointRCNNFeatureExtractor") 11 | class KeypointRCNNFeatureExtractor(nn.Module): 12 | def __init__(self, cfg, in_channels): 13 | super(KeypointRCNNFeatureExtractor, self).__init__() 14 | 15 | resolution = cfg.MODEL.ROI_KEYPOINT_HEAD.POOLER_RESOLUTION 16 | scales = cfg.MODEL.ROI_KEYPOINT_HEAD.POOLER_SCALES 17 | sampling_ratio = cfg.MODEL.ROI_KEYPOINT_HEAD.POOLER_SAMPLING_RATIO 18 | pooler = Pooler( 19 | output_size=(resolution, resolution), 20 | scales=scales, 21 | sampling_ratio=sampling_ratio, 22 | ) 23 | self.pooler = pooler 24 | 25 | input_features = in_channels 26 | layers = cfg.MODEL.ROI_KEYPOINT_HEAD.CONV_LAYERS 27 | next_feature = input_features 28 | self.blocks = [] 29 | for layer_idx, layer_features in enumerate(layers, 1): 30 | layer_name = "conv_fcn{}".format(layer_idx) 31 | module = Conv2d(next_feature, layer_features, 3, stride=1, padding=1) 32 | nn.init.kaiming_normal_(module.weight, mode="fan_out", nonlinearity="relu") 33 | nn.init.constant_(module.bias, 0) 34 | self.add_module(layer_name, module) 35 | next_feature = layer_features 36 | self.blocks.append(layer_name) 37 | self.out_channels = layer_features 38 | 39 | def forward(self, x, proposals): 40 | x = self.pooler(x, proposals) 41 | for layer_name in self.blocks: 42 | x = F.relu(getattr(self, layer_name)(x)) 43 | return x 44 | 45 | 46 | def make_roi_keypoint_feature_extractor(cfg, in_channels): 47 | func = registry.ROI_KEYPOINT_FEATURE_EXTRACTORS[ 48 | cfg.MODEL.ROI_KEYPOINT_HEAD.FEATURE_EXTRACTOR 49 | ] 50 | return func(cfg, in_channels) 51 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/metric_logger.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from collections import defaultdict 3 | from collections import deque 4 | 5 | import torch 6 | 7 | 8 | class SmoothedValue(object): 9 | """Track a series of values and provide access to smoothed values over a 10 | window or the global series average. 11 | """ 12 | 13 | def __init__(self, window_size=20): 14 | self.deque = deque(maxlen=window_size) 15 | self.series = [] 16 | self.total = 0.0 17 | self.count = 0 18 | 19 | def update(self, value): 20 | self.deque.append(value) 21 | self.series.append(value) 22 | self.count += 1 23 | self.total += value 24 | 25 | @property 26 | def median(self): 27 | d = torch.tensor(list(self.deque)) 28 | return d.median().item() 29 | 30 | @property 31 | def avg(self): 32 | d = torch.tensor(list(self.deque)) 33 | return d.mean().item() 34 | 35 | @property 36 | def global_avg(self): 37 | return self.total / self.count 38 | 39 | 40 | class MetricLogger(object): 41 | def __init__(self, delimiter="\t"): 42 | self.meters = defaultdict(SmoothedValue) 43 | self.delimiter = delimiter 44 | 45 | def update(self, **kwargs): 46 | for k, v in kwargs.items(): 47 | if isinstance(v, torch.Tensor): 48 | v = v.item() 49 | assert isinstance(v, (float, int)) 50 | self.meters[k].update(v) 51 | 52 | def __getattr__(self, attr): 53 | if attr in self.meters: 54 | return self.meters[attr] 55 | if attr in self.__dict__: 56 | return self.__dict__[attr] 57 | raise AttributeError("'{}' object has no attribute '{}'".format( 58 | type(self).__name__, attr)) 59 | 60 | def __str__(self): 61 | loss_str = [] 62 | for name, meter in self.meters.items(): 63 | loss_str.append( 64 | "{}: {:.4f} ({:.4f})".format(name, meter.median, meter.global_avg) 65 | ) 66 | return self.delimiter.join(loss_str) 67 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/roi_pool.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 maskrcnn_benchmark import _C 9 | 10 | from apex import amp 11 | 12 | class _ROIPool(Function): 13 | @staticmethod 14 | def forward(ctx, input, roi, output_size, spatial_scale): 15 | ctx.output_size = _pair(output_size) 16 | ctx.spatial_scale = spatial_scale 17 | ctx.input_shape = input.size() 18 | output, argmax = _C.roi_pool_forward( 19 | input, roi, spatial_scale, output_size[0], output_size[1] 20 | ) 21 | ctx.save_for_backward(input, roi, argmax) 22 | return output 23 | 24 | @staticmethod 25 | @once_differentiable 26 | def backward(ctx, grad_output): 27 | input, rois, argmax = ctx.saved_tensors 28 | output_size = ctx.output_size 29 | spatial_scale = ctx.spatial_scale 30 | bs, ch, h, w = ctx.input_shape 31 | grad_input = _C.roi_pool_backward( 32 | grad_output, 33 | input, 34 | rois, 35 | argmax, 36 | spatial_scale, 37 | output_size[0], 38 | output_size[1], 39 | bs, 40 | ch, 41 | h, 42 | w, 43 | ) 44 | return grad_input, None, None, None 45 | 46 | 47 | roi_pool = _ROIPool.apply 48 | 49 | 50 | class ROIPool(nn.Module): 51 | def __init__(self, output_size, spatial_scale): 52 | super(ROIPool, self).__init__() 53 | self.output_size = output_size 54 | self.spatial_scale = spatial_scale 55 | 56 | @amp.float_function 57 | def forward(self, input, rois): 58 | return roi_pool(input, rois, self.output_size, self.spatial_scale) 59 | 60 | def __repr__(self): 61 | tmpstr = self.__class__.__name__ + "(" 62 | tmpstr += "output_size=" + str(self.output_size) 63 | tmpstr += ", spatial_scale=" + str(self.spatial_scale) 64 | tmpstr += ")" 65 | return tmpstr 66 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/keypoint_head.py: -------------------------------------------------------------------------------- 1 | import torch 2 | 3 | from .roi_keypoint_feature_extractors import make_roi_keypoint_feature_extractor 4 | from .roi_keypoint_predictors import make_roi_keypoint_predictor 5 | from .inference import make_roi_keypoint_post_processor 6 | from .loss import make_roi_keypoint_loss_evaluator 7 | 8 | 9 | class ROIKeypointHead(torch.nn.Module): 10 | def __init__(self, cfg, in_channels): 11 | super(ROIKeypointHead, self).__init__() 12 | self.cfg = cfg.clone() 13 | self.feature_extractor = make_roi_keypoint_feature_extractor(cfg, in_channels) 14 | self.predictor = make_roi_keypoint_predictor( 15 | cfg, self.feature_extractor.out_channels) 16 | self.post_processor = make_roi_keypoint_post_processor(cfg) 17 | self.loss_evaluator = make_roi_keypoint_loss_evaluator(cfg) 18 | 19 | def forward(self, features, proposals, targets=None): 20 | """ 21 | Arguments: 22 | features (list[Tensor]): feature-maps from possibly several levels 23 | proposals (list[BoxList]): proposal boxes 24 | targets (list[BoxList], optional): the ground-truth targets. 25 | 26 | Returns: 27 | x (Tensor): the result of the feature extractor 28 | proposals (list[BoxList]): during training, the original proposals 29 | are returned. During testing, the predicted boxlists are returned 30 | with the `mask` field set 31 | losses (dict[Tensor]): During training, returns the losses for the 32 | head. During testing, returns an empty dict. 33 | """ 34 | if self.training: 35 | with torch.no_grad(): 36 | proposals = self.loss_evaluator.subsample(proposals, targets) 37 | 38 | x = self.feature_extractor(features, proposals) 39 | kp_logits = self.predictor(x) 40 | 41 | if not self.training: 42 | result = self.post_processor(kp_logits, proposals) 43 | return x, result, {} 44 | 45 | loss_kp = self.loss_evaluator(proposals, kp_logits) 46 | 47 | return x, proposals, dict(loss_kp=loss_kp) 48 | 49 | 50 | def build_roi_keypoint_head(cfg, in_channels): 51 | return ROIKeypointHead(cfg, in_channels) 52 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/setup.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #!/usr/bin/env python 3 | 4 | import glob 5 | import os 6 | 7 | import torch 8 | from setuptools import find_packages 9 | from setuptools import setup 10 | from torch.utils.cpp_extension import CUDA_HOME 11 | from torch.utils.cpp_extension import CppExtension 12 | from torch.utils.cpp_extension import CUDAExtension 13 | 14 | requirements = ["torch", "torchvision"] 15 | 16 | 17 | def get_extensions(): 18 | this_dir = os.path.dirname(os.path.abspath(__file__)) 19 | extensions_dir = os.path.join(this_dir, "csrc") 20 | 21 | main_file = glob.glob(os.path.join(extensions_dir, "*.cpp")) 22 | source_cpu = glob.glob(os.path.join(extensions_dir, "cpu", "*.cpp")) 23 | source_cuda = glob.glob(os.path.join(extensions_dir, "cuda", "*.cu")) 24 | 25 | sources = main_file + source_cpu 26 | extension = CppExtension 27 | 28 | extra_compile_args = {"cxx": []} 29 | define_macros = [] 30 | 31 | if (torch.cuda.is_available() and CUDA_HOME is not None) or os.getenv("FORCE_CUDA", "0") == "1": 32 | extension = CUDAExtension 33 | sources += source_cuda 34 | define_macros += [("WITH_CUDA", None)] 35 | extra_compile_args["nvcc"] = [ 36 | "-DCUDA_HAS_FP16=1", 37 | "-D__CUDA_NO_HALF_OPERATORS__", 38 | "-D__CUDA_NO_HALF_CONVERSIONS__", 39 | "-D__CUDA_NO_HALF2_OPERATORS__", 40 | ] 41 | 42 | sources = [os.path.join(extensions_dir, s) for s in sources] 43 | 44 | include_dirs = [extensions_dir] 45 | 46 | ext_modules = [ 47 | extension( 48 | "_C", 49 | sources, 50 | include_dirs=include_dirs, 51 | define_macros=define_macros, 52 | extra_compile_args=extra_compile_args, 53 | ) 54 | ] 55 | 56 | return ext_modules 57 | 58 | 59 | setup( 60 | name="maskrcnn_benchmark", 61 | version="0.1", 62 | author="fmassa", 63 | url="https://github.com/facebookresearch/maskrcnn-benchmark", 64 | description="object detection in pytorch", 65 | packages=find_packages(exclude=("configs", "tests",)), 66 | # install_requires=requirements, 67 | ext_modules=get_extensions(), 68 | cmdclass={"build_ext": torch.utils.cpp_extension.BuildExtension}, 69 | ) 70 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/roi_align.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 maskrcnn_benchmark import _C 9 | 10 | from apex import amp 11 | 12 | class _ROIAlign(Function): 13 | @staticmethod 14 | def forward(ctx, input, roi, output_size, spatial_scale, sampling_ratio): 15 | ctx.save_for_backward(roi) 16 | ctx.output_size = _pair(output_size) 17 | ctx.spatial_scale = spatial_scale 18 | ctx.sampling_ratio = sampling_ratio 19 | ctx.input_shape = input.size() 20 | output = _C.roi_align_forward( 21 | input, roi, spatial_scale, output_size[0], output_size[1], sampling_ratio 22 | ) 23 | return output 24 | 25 | @staticmethod 26 | @once_differentiable 27 | def backward(ctx, grad_output): 28 | rois, = ctx.saved_tensors 29 | output_size = ctx.output_size 30 | spatial_scale = ctx.spatial_scale 31 | sampling_ratio = ctx.sampling_ratio 32 | bs, ch, h, w = ctx.input_shape 33 | grad_input = _C.roi_align_backward( 34 | grad_output, 35 | rois, 36 | spatial_scale, 37 | output_size[0], 38 | output_size[1], 39 | bs, 40 | ch, 41 | h, 42 | w, 43 | sampling_ratio, 44 | ) 45 | return grad_input, None, None, None, None 46 | 47 | 48 | roi_align = _ROIAlign.apply 49 | 50 | class ROIAlign(nn.Module): 51 | def __init__(self, output_size, spatial_scale, sampling_ratio): 52 | super(ROIAlign, self).__init__() 53 | self.output_size = output_size 54 | self.spatial_scale = spatial_scale 55 | self.sampling_ratio = sampling_ratio 56 | 57 | @amp.float_function 58 | def forward(self, input, rois): 59 | return roi_align( 60 | input, rois, self.output_size, self.spatial_scale, self.sampling_ratio 61 | ) 62 | 63 | def __repr__(self): 64 | tmpstr = self.__class__.__name__ + "(" 65 | tmpstr += "output_size=" + str(self.output_size) 66 | tmpstr += ", spatial_scale=" + str(self.spatial_scale) 67 | tmpstr += ", sampling_ratio=" + str(self.sampling_ratio) 68 | tmpstr += ")" 69 | return tmpstr 70 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/roi_mask_predictors.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from torch import nn 3 | from torch.nn import functional as F 4 | 5 | from maskrcnn_benchmark.layers import Conv2d 6 | from maskrcnn_benchmark.layers import ConvTranspose2d 7 | from maskrcnn_benchmark.modeling import registry 8 | 9 | 10 | @registry.ROI_MASK_PREDICTOR.register("MaskRCNNC4Predictor") 11 | class MaskRCNNC4Predictor(nn.Module): 12 | def __init__(self, cfg, in_channels): 13 | super(MaskRCNNC4Predictor, self).__init__() 14 | num_classes = cfg.MODEL.ROI_BOX_HEAD.NUM_CLASSES 15 | dim_reduced = cfg.MODEL.ROI_MASK_HEAD.CONV_LAYERS[-1] 16 | num_inputs = in_channels 17 | 18 | self.conv5_mask = ConvTranspose2d(num_inputs, dim_reduced, 2, 2, 0) 19 | self.mask_fcn_logits = Conv2d(dim_reduced, num_classes, 1, 1, 0) 20 | 21 | for name, param in self.named_parameters(): 22 | if "bias" in name: 23 | nn.init.constant_(param, 0) 24 | elif "weight" in name: 25 | # Caffe2 implementation uses MSRAFill, which in fact 26 | # corresponds to kaiming_normal_ in PyTorch 27 | nn.init.kaiming_normal_(param, mode="fan_out", nonlinearity="relu") 28 | 29 | def forward(self, x): 30 | x = F.relu(self.conv5_mask(x)) 31 | return self.mask_fcn_logits(x) 32 | 33 | 34 | @registry.ROI_MASK_PREDICTOR.register("MaskRCNNConv1x1Predictor") 35 | class MaskRCNNConv1x1Predictor(nn.Module): 36 | def __init__(self, cfg, in_channels): 37 | super(MaskRCNNConv1x1Predictor, self).__init__() 38 | num_classes = cfg.MODEL.ROI_BOX_HEAD.NUM_CLASSES 39 | num_inputs = in_channels 40 | 41 | self.mask_fcn_logits = Conv2d(num_inputs, num_classes, 1, 1, 0) 42 | 43 | for name, param in self.named_parameters(): 44 | if "bias" in name: 45 | nn.init.constant_(param, 0) 46 | elif "weight" in name: 47 | # Caffe2 implementation uses MSRAFill, which in fact 48 | # corresponds to kaiming_normal_ in PyTorch 49 | nn.init.kaiming_normal_(param, mode="fan_out", nonlinearity="relu") 50 | 51 | def forward(self, x): 52 | return self.mask_fcn_logits(x) 53 | 54 | 55 | def make_roi_mask_predictor(cfg, in_channels): 56 | func = registry.ROI_MASK_PREDICTOR[cfg.MODEL.ROI_MASK_HEAD.PREDICTOR] 57 | return func(cfg, in_channels) 58 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/detector/generalized_rcnn.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | """ 3 | Implements the Generalized R-CNN framework 4 | """ 5 | 6 | import torch 7 | from torch import nn 8 | 9 | from maskrcnn_benchmark.structures.image_list import to_image_list 10 | 11 | from ..backbone import build_backbone 12 | from ..rpn.rpn import build_rpn 13 | from ..roi_heads.roi_heads import build_roi_heads 14 | 15 | 16 | class GeneralizedRCNN(nn.Module): 17 | """ 18 | Main class for Generalized R-CNN. Currently supports boxes and masks. 19 | It consists of three main parts: 20 | - backbone 21 | - rpn 22 | - heads: takes the features + the proposals from the RPN and computes 23 | detections / masks from it. 24 | """ 25 | 26 | def __init__(self, cfg): 27 | super(GeneralizedRCNN, self).__init__() 28 | 29 | self.backbone = build_backbone(cfg) 30 | self.rpn = build_rpn(cfg, self.backbone.out_channels) 31 | self.roi_heads = build_roi_heads(cfg, self.backbone.out_channels) 32 | 33 | def forward(self, images, targets=None): 34 | """ 35 | Arguments: 36 | images (list[Tensor] or ImageList): images to be processed 37 | targets (list[BoxList]): ground-truth boxes present in the image (optional) 38 | 39 | Returns: 40 | result (list[BoxList] or dict[Tensor]): the output from the model. 41 | During training, it returns a dict[Tensor] which contains the losses. 42 | During testing, it returns list[BoxList] contains additional fields 43 | like `scores`, `labels` and `mask` (for Mask R-CNN models). 44 | 45 | """ 46 | if self.training and targets is None: 47 | raise ValueError("In training mode, targets should be passed") 48 | images = to_image_list(images) 49 | features = self.backbone(images.tensors) 50 | proposals, proposal_losses = self.rpn(images, features, targets) 51 | if self.roi_heads: 52 | x, result, detector_losses = self.roi_heads(features, proposals, targets) 53 | else: 54 | # RPN-only models don't have roi_heads 55 | x = features 56 | result = proposals 57 | detector_losses = {} 58 | 59 | if self.training: 60 | losses = {} 61 | losses.update(detector_losses) 62 | losses.update(proposal_losses) 63 | return losses 64 | 65 | return result 66 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/box_head/roi_box_predictors.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from maskrcnn_benchmark.modeling import registry 3 | from torch import nn 4 | 5 | 6 | @registry.ROI_BOX_PREDICTOR.register("FastRCNNPredictor") 7 | class FastRCNNPredictor(nn.Module): 8 | def __init__(self, config, in_channels): 9 | super(FastRCNNPredictor, self).__init__() 10 | assert in_channels is not None 11 | 12 | num_inputs = in_channels 13 | 14 | num_classes = config.MODEL.ROI_BOX_HEAD.NUM_CLASSES 15 | self.avgpool = nn.AdaptiveAvgPool2d(1) 16 | self.cls_score = nn.Linear(num_inputs, num_classes) 17 | num_bbox_reg_classes = 2 if config.MODEL.CLS_AGNOSTIC_BBOX_REG else num_classes 18 | self.bbox_pred = nn.Linear(num_inputs, num_bbox_reg_classes * 4) 19 | 20 | nn.init.normal_(self.cls_score.weight, mean=0, std=0.01) 21 | nn.init.constant_(self.cls_score.bias, 0) 22 | 23 | nn.init.normal_(self.bbox_pred.weight, mean=0, std=0.001) 24 | nn.init.constant_(self.bbox_pred.bias, 0) 25 | 26 | def forward(self, x): 27 | x = self.avgpool(x) 28 | x = x.view(x.size(0), -1) 29 | cls_logit = self.cls_score(x) 30 | bbox_pred = self.bbox_pred(x) 31 | return cls_logit, bbox_pred 32 | 33 | 34 | @registry.ROI_BOX_PREDICTOR.register("FPNPredictor") 35 | class FPNPredictor(nn.Module): 36 | def __init__(self, cfg, in_channels): 37 | super(FPNPredictor, self).__init__() 38 | num_classes = cfg.MODEL.ROI_BOX_HEAD.NUM_CLASSES 39 | representation_size = in_channels 40 | 41 | self.cls_score = nn.Linear(representation_size, num_classes) 42 | num_bbox_reg_classes = 2 if cfg.MODEL.CLS_AGNOSTIC_BBOX_REG else num_classes 43 | self.bbox_pred = nn.Linear(representation_size, num_bbox_reg_classes * 4) 44 | 45 | nn.init.normal_(self.cls_score.weight, std=0.01) 46 | nn.init.normal_(self.bbox_pred.weight, std=0.001) 47 | for l in [self.cls_score, self.bbox_pred]: 48 | nn.init.constant_(l.bias, 0) 49 | 50 | def forward(self, x): 51 | if x.ndimension() == 4: 52 | assert list(x.shape[2:]) == [1, 1] 53 | x = x.view(x.size(0), -1) 54 | scores = self.cls_score(x) 55 | bbox_deltas = self.bbox_pred(x) 56 | 57 | return scores, bbox_deltas 58 | 59 | 60 | def make_roi_box_predictor(cfg, in_channels): 61 | func = registry.ROI_BOX_PREDICTOR[cfg.MODEL.ROI_BOX_HEAD.PREDICTOR] 62 | return func(cfg, in_channels) 63 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/sigmoid_focal_loss.py: -------------------------------------------------------------------------------- 1 | import torch 2 | from torch import nn 3 | from torch.autograd import Function 4 | from torch.autograd.function import once_differentiable 5 | 6 | from maskrcnn_benchmark import _C 7 | 8 | # TODO: Use JIT to replace CUDA implementation in the future. 9 | class _SigmoidFocalLoss(Function): 10 | @staticmethod 11 | def forward(ctx, logits, targets, gamma, alpha): 12 | ctx.save_for_backward(logits, targets) 13 | num_classes = logits.shape[1] 14 | ctx.num_classes = num_classes 15 | ctx.gamma = gamma 16 | ctx.alpha = alpha 17 | 18 | losses = _C.sigmoid_focalloss_forward( 19 | logits, targets, num_classes, gamma, alpha 20 | ) 21 | return losses 22 | 23 | @staticmethod 24 | @once_differentiable 25 | def backward(ctx, d_loss): 26 | logits, targets = ctx.saved_tensors 27 | num_classes = ctx.num_classes 28 | gamma = ctx.gamma 29 | alpha = ctx.alpha 30 | d_loss = d_loss.contiguous() 31 | d_logits = _C.sigmoid_focalloss_backward( 32 | logits, targets, d_loss, num_classes, gamma, alpha 33 | ) 34 | return d_logits, None, None, None, None 35 | 36 | 37 | sigmoid_focal_loss_cuda = _SigmoidFocalLoss.apply 38 | 39 | 40 | def sigmoid_focal_loss_cpu(logits, targets, gamma, alpha): 41 | num_classes = logits.shape[1] 42 | gamma = gamma[0] 43 | alpha = alpha[0] 44 | dtype = targets.dtype 45 | device = targets.device 46 | class_range = torch.arange(1, num_classes+1, dtype=dtype, device=device).unsqueeze(0) 47 | 48 | t = targets.unsqueeze(1) 49 | p = torch.sigmoid(logits) 50 | term1 = (1 - p) ** gamma * torch.log(p) 51 | term2 = p ** gamma * torch.log(1 - p) 52 | return -(t == class_range).float() * term1 * alpha - ((t != class_range) * (t >= 0)).float() * term2 * (1 - alpha) 53 | 54 | 55 | class SigmoidFocalLoss(nn.Module): 56 | def __init__(self, gamma, alpha): 57 | super(SigmoidFocalLoss, self).__init__() 58 | self.gamma = gamma 59 | self.alpha = alpha 60 | 61 | def forward(self, logits, targets): 62 | device = logits.device 63 | if logits.is_cuda: 64 | loss_func = sigmoid_focal_loss_cuda 65 | else: 66 | loss_func = sigmoid_focal_loss_cpu 67 | 68 | loss = loss_func(logits, targets, self.gamma, self.alpha) 69 | return loss.sum() 70 | 71 | def __repr__(self): 72 | tmpstr = self.__class__.__name__ + "(" 73 | tmpstr += "gamma=" + str(self.gamma) 74 | tmpstr += ", alpha=" + str(self.alpha) 75 | tmpstr += ")" 76 | return tmpstr 77 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/cpu/nms_cpu.cpp: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #include "cpu/vision.h" 3 | 4 | 5 | template 6 | at::Tensor nms_cpu_kernel(const at::Tensor& dets, 7 | const at::Tensor& scores, 8 | const float threshold) { 9 | AT_ASSERTM(!dets.type().is_cuda(), "dets must be a CPU tensor"); 10 | AT_ASSERTM(!scores.type().is_cuda(), "scores must be a CPU tensor"); 11 | AT_ASSERTM(dets.type() == scores.type(), "dets should have the same type as scores"); 12 | 13 | if (dets.numel() == 0) { 14 | return at::empty({0}, dets.options().dtype(at::kLong).device(at::kCPU)); 15 | } 16 | 17 | auto x1_t = dets.select(1, 0).contiguous(); 18 | auto y1_t = dets.select(1, 1).contiguous(); 19 | auto x2_t = dets.select(1, 2).contiguous(); 20 | auto y2_t = dets.select(1, 3).contiguous(); 21 | 22 | at::Tensor areas_t = (x2_t - x1_t + 1) * (y2_t - y1_t + 1); 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).device(at::kCPU)); 28 | 29 | auto suppressed = suppressed_t.data(); 30 | auto order = order_t.data(); 31 | auto x1 = x1_t.data(); 32 | auto y1 = y1_t.data(); 33 | auto x2 = x2_t.data(); 34 | auto y2 = y2_t.data(); 35 | auto areas = areas_t.data(); 36 | 37 | for (int64_t _i = 0; _i < ndets; _i++) { 38 | auto i = order[_i]; 39 | if (suppressed[i] == 1) 40 | continue; 41 | auto ix1 = x1[i]; 42 | auto iy1 = y1[i]; 43 | auto ix2 = x2[i]; 44 | auto iy2 = y2[i]; 45 | auto iarea = areas[i]; 46 | 47 | for (int64_t _j = _i + 1; _j < ndets; _j++) { 48 | auto j = order[_j]; 49 | if (suppressed[j] == 1) 50 | continue; 51 | auto xx1 = std::max(ix1, x1[j]); 52 | auto yy1 = std::max(iy1, y1[j]); 53 | auto xx2 = std::min(ix2, x2[j]); 54 | auto yy2 = std::min(iy2, y2[j]); 55 | 56 | auto w = std::max(static_cast(0), xx2 - xx1 + 1); 57 | auto h = std::max(static_cast(0), yy2 - yy1 + 1); 58 | auto inter = w * h; 59 | auto ovr = inter / (iarea + areas[j] - inter); 60 | if (ovr >= threshold) 61 | suppressed[j] = 1; 62 | } 63 | } 64 | return at::nonzero(suppressed_t == 0).squeeze(1); 65 | } 66 | 67 | at::Tensor nms_cpu(const at::Tensor& dets, 68 | const at::Tensor& scores, 69 | const float threshold) { 70 | at::Tensor result; 71 | AT_DISPATCH_FLOATING_TYPES(dets.type(), "nms", [&] { 72 | result = nms_cpu_kernel(dets, scores, threshold); 73 | }); 74 | return result; 75 | } 76 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/structures/image_list.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from __future__ import division 3 | 4 | import torch 5 | 6 | 7 | class ImageList(object): 8 | """ 9 | Structure that holds a list of images (of possibly 10 | varying sizes) as a single tensor. 11 | This works by padding the images to the same size, 12 | and storing in a field the original sizes of each image 13 | """ 14 | 15 | def __init__(self, tensors, image_sizes): 16 | """ 17 | Arguments: 18 | tensors (tensor) 19 | image_sizes (list[tuple[int, int]]) 20 | """ 21 | self.tensors = tensors 22 | self.image_sizes = image_sizes 23 | 24 | def to(self, *args, **kwargs): 25 | cast_tensor = self.tensors.to(*args, **kwargs) 26 | return ImageList(cast_tensor, self.image_sizes) 27 | 28 | 29 | def to_image_list(tensors, size_divisible=0): 30 | """ 31 | tensors can be an ImageList, a torch.Tensor or 32 | an iterable of Tensors. It can't be a numpy array. 33 | When tensors is an iterable of Tensors, it pads 34 | the Tensors with zeros so that they have the same 35 | shape 36 | """ 37 | if isinstance(tensors, torch.Tensor) and size_divisible > 0: 38 | tensors = [tensors] 39 | 40 | if isinstance(tensors, ImageList): 41 | return tensors 42 | elif isinstance(tensors, torch.Tensor): 43 | # single tensor shape can be inferred 44 | if tensors.dim() == 3: 45 | tensors = tensors[None] 46 | assert tensors.dim() == 4 47 | image_sizes = [tensor.shape[-2:] for tensor in tensors] 48 | return ImageList(tensors, image_sizes) 49 | elif isinstance(tensors, (tuple, list)): 50 | max_size = tuple(max(s) for s in zip(*[img.shape for img in tensors])) 51 | 52 | # TODO Ideally, just remove this and let me model handle arbitrary 53 | # input sizs 54 | if size_divisible > 0: 55 | import math 56 | 57 | stride = size_divisible 58 | max_size = list(max_size) 59 | max_size[1] = int(math.ceil(max_size[1] / stride) * stride) 60 | max_size[2] = int(math.ceil(max_size[2] / stride) * stride) 61 | max_size = tuple(max_size) 62 | 63 | batch_shape = (len(tensors),) + max_size 64 | batched_imgs = tensors[0].new(*batch_shape).zero_() 65 | for img, pad_img in zip(tensors, batched_imgs): 66 | pad_img[: img.shape[0], : img.shape[1], : img.shape[2]].copy_(img) 67 | 68 | image_sizes = [im.shape[-2:] for im in tensors] 69 | 70 | return ImageList(batched_imgs, image_sizes) 71 | else: 72 | raise TypeError("Unsupported type for to_image_list: {}".format(type(tensors))) 73 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/roi_mask_feature_extractors.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from torch import nn 3 | from torch.nn import functional as F 4 | 5 | from ..box_head.roi_box_feature_extractors import ResNet50Conv5ROIFeatureExtractor 6 | from maskrcnn_benchmark.modeling import registry 7 | from maskrcnn_benchmark.modeling.poolers import Pooler 8 | from maskrcnn_benchmark.modeling.make_layers import make_conv3x3 9 | 10 | 11 | registry.ROI_MASK_FEATURE_EXTRACTORS.register( 12 | "ResNet50Conv5ROIFeatureExtractor", ResNet50Conv5ROIFeatureExtractor 13 | ) 14 | 15 | 16 | @registry.ROI_MASK_FEATURE_EXTRACTORS.register("MaskRCNNFPNFeatureExtractor") 17 | class MaskRCNNFPNFeatureExtractor(nn.Module): 18 | """ 19 | Heads for FPN for classification 20 | """ 21 | 22 | def __init__(self, cfg, in_channels): 23 | """ 24 | Arguments: 25 | num_classes (int): number of output classes 26 | input_size (int): number of channels of the input once it's flattened 27 | representation_size (int): size of the intermediate representation 28 | """ 29 | super(MaskRCNNFPNFeatureExtractor, self).__init__() 30 | 31 | resolution = cfg.MODEL.ROI_MASK_HEAD.POOLER_RESOLUTION 32 | scales = cfg.MODEL.ROI_MASK_HEAD.POOLER_SCALES 33 | sampling_ratio = cfg.MODEL.ROI_MASK_HEAD.POOLER_SAMPLING_RATIO 34 | pooler = Pooler( 35 | output_size=(resolution, resolution), 36 | scales=scales, 37 | sampling_ratio=sampling_ratio, 38 | ) 39 | input_size = in_channels 40 | self.pooler = pooler 41 | 42 | use_gn = cfg.MODEL.ROI_MASK_HEAD.USE_GN 43 | layers = cfg.MODEL.ROI_MASK_HEAD.CONV_LAYERS 44 | dilation = cfg.MODEL.ROI_MASK_HEAD.DILATION 45 | 46 | next_feature = input_size 47 | self.blocks = [] 48 | for layer_idx, layer_features in enumerate(layers, 1): 49 | layer_name = "mask_fcn{}".format(layer_idx) 50 | module = make_conv3x3( 51 | next_feature, layer_features, 52 | dilation=dilation, stride=1, use_gn=use_gn 53 | ) 54 | self.add_module(layer_name, module) 55 | next_feature = layer_features 56 | self.blocks.append(layer_name) 57 | self.out_channels = layer_features 58 | 59 | def forward(self, x, proposals): 60 | x = self.pooler(x, proposals) 61 | 62 | for layer_name in self.blocks: 63 | x = F.relu(getattr(self, layer_name)(x)) 64 | 65 | return x 66 | 67 | 68 | def make_roi_mask_feature_extractor(cfg, in_channels): 69 | func = registry.ROI_MASK_FEATURE_EXTRACTORS[ 70 | cfg.MODEL.ROI_MASK_HEAD.FEATURE_EXTRACTOR 71 | ] 72 | return func(cfg, in_channels) 73 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/samplers/distributed.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | # Code is copy-pasted exactly as in torch.utils.data.distributed. 3 | # FIXME remove this once c10d fixes the bug it has 4 | import math 5 | import torch 6 | import torch.distributed as dist 7 | from torch.utils.data.sampler import Sampler 8 | 9 | 10 | class DistributedSampler(Sampler): 11 | """Sampler that restricts data loading to a subset of the dataset. 12 | It is especially useful in conjunction with 13 | :class:`torch.nn.parallel.DistributedDataParallel`. In such case, each 14 | process can pass a DistributedSampler instance as a DataLoader sampler, 15 | and load a subset of the original dataset that is exclusive to it. 16 | .. note:: 17 | Dataset is assumed to be of constant size. 18 | Arguments: 19 | dataset: Dataset used for sampling. 20 | num_replicas (optional): Number of processes participating in 21 | distributed training. 22 | rank (optional): Rank of the current process within num_replicas. 23 | """ 24 | 25 | def __init__(self, dataset, num_replicas=None, rank=None, shuffle=True): 26 | if num_replicas is None: 27 | if not dist.is_available(): 28 | raise RuntimeError("Requires distributed package to be available") 29 | num_replicas = dist.get_world_size() 30 | if rank is None: 31 | if not dist.is_available(): 32 | raise RuntimeError("Requires distributed package to be available") 33 | rank = dist.get_rank() 34 | self.dataset = dataset 35 | self.num_replicas = num_replicas 36 | self.rank = rank 37 | self.epoch = 0 38 | self.num_samples = int(math.ceil(len(self.dataset) * 1.0 / self.num_replicas)) 39 | self.total_size = self.num_samples * self.num_replicas 40 | self.shuffle = shuffle 41 | 42 | def __iter__(self): 43 | if self.shuffle: 44 | # deterministically shuffle based on epoch 45 | g = torch.Generator() 46 | g.manual_seed(self.epoch) 47 | indices = torch.randperm(len(self.dataset), generator=g).tolist() 48 | else: 49 | indices = torch.arange(len(self.dataset)).tolist() 50 | 51 | # add extra samples to make it evenly divisible 52 | indices += indices[: (self.total_size - len(indices))] 53 | assert len(indices) == self.total_size 54 | 55 | # subsample 56 | offset = self.num_samples * self.rank 57 | indices = indices[offset : offset + self.num_samples] 58 | assert len(indices) == self.num_samples 59 | 60 | return iter(indices) 61 | 62 | def __len__(self): 63 | return self.num_samples 64 | 65 | def set_epoch(self, epoch): 66 | self.epoch = epoch 67 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/layers/dcn/deform_pool_func.py: -------------------------------------------------------------------------------- 1 | import torch 2 | from torch.autograd import Function 3 | from torch.autograd.function import once_differentiable 4 | 5 | from maskrcnn_benchmark import _C 6 | 7 | 8 | class DeformRoIPoolingFunction(Function): 9 | 10 | @staticmethod 11 | def forward( 12 | ctx, 13 | data, 14 | rois, 15 | offset, 16 | spatial_scale, 17 | out_size, 18 | out_channels, 19 | no_trans, 20 | group_size=1, 21 | part_size=None, 22 | sample_per_part=4, 23 | trans_std=.0 24 | ): 25 | ctx.spatial_scale = spatial_scale 26 | ctx.out_size = out_size 27 | ctx.out_channels = out_channels 28 | ctx.no_trans = no_trans 29 | ctx.group_size = group_size 30 | ctx.part_size = out_size if part_size is None else part_size 31 | ctx.sample_per_part = sample_per_part 32 | ctx.trans_std = trans_std 33 | 34 | assert 0.0 <= ctx.trans_std <= 1.0 35 | if not data.is_cuda: 36 | raise NotImplementedError 37 | 38 | n = rois.shape[0] 39 | output = data.new_empty(n, out_channels, out_size, out_size) 40 | output_count = data.new_empty(n, out_channels, out_size, out_size) 41 | _C.deform_psroi_pooling_forward( 42 | data, 43 | rois, 44 | offset, 45 | output, 46 | output_count, 47 | ctx.no_trans, 48 | ctx.spatial_scale, 49 | ctx.out_channels, 50 | ctx.group_size, 51 | ctx.out_size, 52 | ctx.part_size, 53 | ctx.sample_per_part, 54 | ctx.trans_std 55 | ) 56 | 57 | if data.requires_grad or rois.requires_grad or offset.requires_grad: 58 | ctx.save_for_backward(data, rois, offset) 59 | ctx.output_count = output_count 60 | 61 | return output 62 | 63 | @staticmethod 64 | @once_differentiable 65 | def backward(ctx, grad_output): 66 | if not grad_output.is_cuda: 67 | raise NotImplementedError 68 | 69 | data, rois, offset = ctx.saved_tensors 70 | output_count = ctx.output_count 71 | grad_input = torch.zeros_like(data) 72 | grad_rois = None 73 | grad_offset = torch.zeros_like(offset) 74 | 75 | _C.deform_psroi_pooling_backward( 76 | grad_output, 77 | data, 78 | rois, 79 | offset, 80 | output_count, 81 | grad_input, 82 | grad_offset, 83 | ctx.no_trans, 84 | ctx.spatial_scale, 85 | ctx.out_channels, 86 | ctx.group_size, 87 | ctx.out_size, 88 | ctx.part_size, 89 | ctx.sample_per_part, 90 | ctx.trans_std 91 | ) 92 | return (grad_input, grad_rois, grad_offset, None, None, None, None, None, None, None, None) 93 | 94 | 95 | deform_roi_pooling = DeformRoIPoolingFunction.apply 96 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/balanced_positive_negative_sampler.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | 4 | 5 | class BalancedPositiveNegativeSampler(object): 6 | """ 7 | This class samples batches, ensuring that they contain a fixed proportion of positives 8 | """ 9 | 10 | def __init__(self, batch_size_per_image, positive_fraction): 11 | """ 12 | Arguments: 13 | batch_size_per_image (int): number of elements to be selected per image 14 | positive_fraction (float): percentage of positive elements per batch 15 | """ 16 | self.batch_size_per_image = batch_size_per_image 17 | self.positive_fraction = positive_fraction 18 | 19 | def __call__(self, matched_idxs): 20 | """ 21 | Arguments: 22 | matched idxs: list of tensors containing -1, 0 or positive values. 23 | Each tensor corresponds to a specific image. 24 | -1 values are ignored, 0 are considered as negatives and > 0 as 25 | positives. 26 | 27 | Returns: 28 | pos_idx (list[tensor]) 29 | neg_idx (list[tensor]) 30 | 31 | Returns two lists of binary masks for each image. 32 | The first list contains the positive elements that were selected, 33 | and the second list the negative example. 34 | """ 35 | pos_idx = [] 36 | neg_idx = [] 37 | for matched_idxs_per_image in matched_idxs: 38 | positive = torch.nonzero(matched_idxs_per_image >= 1).squeeze(1) 39 | negative = torch.nonzero(matched_idxs_per_image == 0).squeeze(1) 40 | 41 | num_pos = int(self.batch_size_per_image * self.positive_fraction) 42 | # protect against not enough positive examples 43 | num_pos = min(positive.numel(), num_pos) 44 | num_neg = self.batch_size_per_image - num_pos 45 | # protect against not enough negative examples 46 | num_neg = min(negative.numel(), num_neg) 47 | 48 | # randomly select positive and negative examples 49 | perm1 = torch.randperm(positive.numel(), device=positive.device)[:num_pos] 50 | perm2 = torch.randperm(negative.numel(), device=negative.device)[:num_neg] 51 | 52 | pos_idx_per_image = positive[perm1] 53 | neg_idx_per_image = negative[perm2] 54 | 55 | # create binary mask from indices 56 | pos_idx_per_image_mask = torch.zeros_like( 57 | matched_idxs_per_image, dtype=torch.uint8 58 | ) 59 | neg_idx_per_image_mask = torch.zeros_like( 60 | matched_idxs_per_image, dtype=torch.uint8 61 | ) 62 | pos_idx_per_image_mask[pos_idx_per_image] = 1 63 | neg_idx_per_image_mask[neg_idx_per_image] = 1 64 | 65 | pos_idx.append(pos_idx_per_image_mask) 66 | neg_idx.append(neg_idx_per_image_mask) 67 | 68 | return pos_idx, neg_idx 69 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/box_head/box_head.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | from torch import nn 4 | 5 | from .roi_box_feature_extractors import make_roi_box_feature_extractor 6 | from .roi_box_predictors import make_roi_box_predictor 7 | from .inference import make_roi_box_post_processor 8 | from .loss import make_roi_box_loss_evaluator 9 | 10 | 11 | class ROIBoxHead(torch.nn.Module): 12 | """ 13 | Generic Box Head class. 14 | """ 15 | 16 | def __init__(self, cfg, in_channels): 17 | super(ROIBoxHead, self).__init__() 18 | self.feature_extractor = make_roi_box_feature_extractor(cfg, in_channels) 19 | self.predictor = make_roi_box_predictor( 20 | cfg, self.feature_extractor.out_channels) 21 | self.post_processor = make_roi_box_post_processor(cfg) 22 | self.loss_evaluator = make_roi_box_loss_evaluator(cfg) 23 | 24 | def forward(self, features, proposals, targets=None): 25 | """ 26 | Arguments: 27 | features (list[Tensor]): feature-maps from possibly several levels 28 | proposals (list[BoxList]): proposal boxes 29 | targets (list[BoxList], optional): the ground-truth targets. 30 | 31 | Returns: 32 | x (Tensor): the result of the feature extractor 33 | proposals (list[BoxList]): during training, the subsampled proposals 34 | are returned. During testing, the predicted boxlists are returned 35 | losses (dict[Tensor]): During training, returns the losses for the 36 | head. During testing, returns an empty dict. 37 | """ 38 | 39 | if self.training: 40 | # Faster R-CNN subsamples during training the proposals with a fixed 41 | # positive / negative ratio 42 | with torch.no_grad(): 43 | proposals = self.loss_evaluator.subsample(proposals, targets) 44 | 45 | # extract features that will be fed to the final classifier. The 46 | # feature_extractor generally corresponds to the pooler + heads 47 | x = self.feature_extractor(features, proposals) 48 | # final classifier that converts the features into predictions 49 | class_logits, box_regression = self.predictor(x) 50 | 51 | if not self.training: 52 | result = self.post_processor((class_logits, box_regression), proposals) 53 | return x, result, {} 54 | 55 | loss_classifier, loss_box_reg = self.loss_evaluator( 56 | [class_logits], [box_regression] 57 | ) 58 | return ( 59 | x, 60 | proposals, 61 | dict(loss_classifier=loss_classifier, loss_box_reg=loss_box_reg), 62 | ) 63 | 64 | 65 | def build_roi_box_head(cfg, in_channels): 66 | """ 67 | Constructs a new box head. 68 | By default, uses ROIBoxHead, but if it turns out not to be enough, just register a new class 69 | and make it a parameter in the config 70 | """ 71 | return ROIBoxHead(cfg, in_channels) 72 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/backbone.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from collections import OrderedDict 3 | 4 | from torch import nn 5 | 6 | from maskrcnn_benchmark.modeling import registry 7 | from maskrcnn_benchmark.modeling.make_layers import conv_with_kaiming_uniform 8 | from . import fpn as fpn_module 9 | from . import resnet 10 | 11 | 12 | @registry.BACKBONES.register("R-50-C4") 13 | @registry.BACKBONES.register("R-50-C5") 14 | @registry.BACKBONES.register("R-101-C4") 15 | @registry.BACKBONES.register("R-101-C5") 16 | def build_resnet_backbone(cfg): 17 | body = resnet.ResNet(cfg) 18 | model = nn.Sequential(OrderedDict([("body", body)])) 19 | model.out_channels = cfg.MODEL.RESNETS.BACKBONE_OUT_CHANNELS 20 | return model 21 | 22 | 23 | @registry.BACKBONES.register("R-50-FPN") 24 | @registry.BACKBONES.register("R-101-FPN") 25 | @registry.BACKBONES.register("R-152-FPN") 26 | def build_resnet_fpn_backbone(cfg): 27 | body = resnet.ResNet(cfg) 28 | in_channels_stage2 = cfg.MODEL.RESNETS.RES2_OUT_CHANNELS 29 | out_channels = cfg.MODEL.RESNETS.BACKBONE_OUT_CHANNELS 30 | fpn = fpn_module.FPN( 31 | in_channels_list=[ 32 | in_channels_stage2, 33 | in_channels_stage2 * 2, 34 | in_channels_stage2 * 4, 35 | in_channels_stage2 * 8, 36 | ], 37 | out_channels=out_channels, 38 | conv_block=conv_with_kaiming_uniform( 39 | cfg.MODEL.FPN.USE_GN, cfg.MODEL.FPN.USE_RELU 40 | ), 41 | top_blocks=fpn_module.LastLevelMaxPool(), 42 | ) 43 | model = nn.Sequential(OrderedDict([("body", body), ("fpn", fpn)])) 44 | model.out_channels = out_channels 45 | return model 46 | 47 | 48 | @registry.BACKBONES.register("R-50-FPN-RETINANET") 49 | @registry.BACKBONES.register("R-101-FPN-RETINANET") 50 | def build_resnet_fpn_p3p7_backbone(cfg): 51 | body = resnet.ResNet(cfg) 52 | in_channels_stage2 = cfg.MODEL.RESNETS.RES2_OUT_CHANNELS 53 | out_channels = cfg.MODEL.RESNETS.BACKBONE_OUT_CHANNELS 54 | in_channels_p6p7 = in_channels_stage2 * 8 if cfg.MODEL.RETINANET.USE_C5 \ 55 | else out_channels 56 | fpn = fpn_module.FPN( 57 | in_channels_list=[ 58 | 0, 59 | in_channels_stage2 * 2, 60 | in_channels_stage2 * 4, 61 | in_channels_stage2 * 8, 62 | ], 63 | out_channels=out_channels, 64 | conv_block=conv_with_kaiming_uniform( 65 | cfg.MODEL.FPN.USE_GN, cfg.MODEL.FPN.USE_RELU 66 | ), 67 | top_blocks=fpn_module.LastLevelP6P7(in_channels_p6p7, out_channels), 68 | ) 69 | model = nn.Sequential(OrderedDict([("body", body), ("fpn", fpn)])) 70 | model.out_channels = out_channels 71 | return model 72 | 73 | 74 | def build_backbone(cfg): 75 | assert cfg.MODEL.BACKBONE.CONV_BODY in registry.BACKBONES, \ 76 | "cfg.MODEL.BACKBONE.CONV_BODY: {} are not registered in registry".format( 77 | cfg.MODEL.BACKBONE.CONV_BODY 78 | ) 79 | return registry.BACKBONES[cfg.MODEL.BACKBONE.CONV_BODY](cfg) 80 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/README.md: -------------------------------------------------------------------------------- 1 | # Setting Up Datasets 2 | This file describes how to perform training on other datasets. 3 | 4 | Only Pascal VOC dataset can be loaded from its original format and be outputted to Pascal style results currently. 5 | 6 | We expect the annotations from other datasets be converted to COCO json format, and 7 | the output will be in COCO-style. (i.e. AP, AP50, AP75, APs, APm, APl for bbox and segm) 8 | 9 | ## Creating Symlinks for PASCAL VOC 10 | 11 | We assume that your symlinked `datasets/voc/VOC` directory has the following structure: 12 | 13 | ``` 14 | VOC 15 | |_ JPEGImages 16 | | |_ .jpg 17 | | |_ ... 18 | | |_ .jpg 19 | |_ Annotations 20 | | |_ pascal_train.json (optional) 21 | | |_ pascal_val.json (optional) 22 | | |_ pascal_test.json (optional) 23 | | |_ .xml 24 | | |_ ... 25 | | |_ .xml 26 | |_ VOCdevkit 27 | ``` 28 | 29 | Create symlinks for `voc/VOC`: 30 | 31 | ``` 32 | cd ~/github/maskrcnn-benchmark 33 | mkdir -p datasets/voc/VOC 34 | ln -s /path/to/VOC /datasets/voc/VOC 35 | ``` 36 | Example configuration files for PASCAL VOC could be found [here](https://github.com/facebookresearch/maskrcnn-benchmark/blob/master/configs/pascal_voc/). 37 | 38 | ### PASCAL VOC Annotations in COCO Format 39 | To output COCO-style evaluation result, PASCAL VOC annotations in COCO json format is required and could be downloaded from [here](https://storage.googleapis.com/coco-dataset/external/PASCAL_VOC.zip) 40 | via http://cocodataset.org/#external. 41 | 42 | ## Creating Symlinks for Cityscapes: 43 | 44 | We assume that your symlinked `datasets/cityscapes` directory has the following structure: 45 | 46 | ``` 47 | cityscapes 48 | |_ images 49 | | |_ .jpg 50 | | |_ ... 51 | | |_ .jpg 52 | |_ annotations 53 | | |_ instanceonly_gtFile_train.json 54 | | |_ ... 55 | |_ raw 56 | |_ gtFine 57 | |_ ... 58 | |_ README.md 59 | ``` 60 | 61 | Create symlinks for `cityscapes`: 62 | 63 | ``` 64 | cd ~/github/maskrcnn-benchmark 65 | mkdir -p datasets/cityscapes 66 | ln -s /path/to/cityscapes datasets/data/cityscapes 67 | ``` 68 | 69 | ### Steps to convert Cityscapes Annotations to COCO Format 70 | 1. Download gtFine_trainvaltest.zip from https://www.cityscapes-dataset.com/downloads/ (login required) 71 | 2. Extract it to /path/to/gtFine_trainvaltest 72 | ``` 73 | cityscapes 74 | |_ gtFine_trainvaltest.zip 75 | |_ gtFine_trainvaltest 76 | |_ gtFine 77 | ``` 78 | 3. Run the below commands to convert the annotations 79 | 80 | ``` 81 | cd ~/github 82 | git clone https://github.com/mcordts/cityscapesScripts.git 83 | cd cityscapesScripts 84 | cp ~/github/maskrcnn-benchmark/tools/cityscapes/instances2dict_with_polygons.py cityscapesscripts/evaluation 85 | python setup.py install 86 | cd ~/github/maskrcnn-benchmark 87 | python tools/cityscapes/convert_cityscapes_to_coco.py --datadir /path/to/cityscapes --outdir /path/to/cityscapes/annotations 88 | ``` 89 | 90 | Example configuration files for Cityscapes could be found [here](https://github.com/facebookresearch/maskrcnn-benchmark/blob/master/configs/cityscapes/). 91 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/model_zoo.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import os 3 | import sys 4 | 5 | try: 6 | from torch.hub import _download_url_to_file 7 | from torch.hub import urlparse 8 | from torch.hub import HASH_REGEX 9 | except ImportError: 10 | from torch.utils.model_zoo import _download_url_to_file 11 | from torch.utils.model_zoo import urlparse 12 | from torch.utils.model_zoo import HASH_REGEX 13 | 14 | from maskrcnn_benchmark.utils.comm import is_main_process 15 | from maskrcnn_benchmark.utils.comm import synchronize 16 | 17 | 18 | # very similar to https://github.com/pytorch/pytorch/blob/master/torch/utils/model_zoo.py 19 | # but with a few improvements and modifications 20 | def cache_url(url, model_dir=None, progress=True): 21 | r"""Loads the Torch serialized object at the given URL. 22 | If the object is already present in `model_dir`, it's deserialized and 23 | returned. The filename part of the URL should follow the naming convention 24 | ``filename-.ext`` where ```` is the first eight or more 25 | digits of the SHA256 hash of the contents of the file. The hash is used to 26 | ensure unique names and to verify the contents of the file. 27 | The default value of `model_dir` is ``$TORCH_HOME/models`` where 28 | ``$TORCH_HOME`` defaults to ``~/.torch``. The default directory can be 29 | overridden with the ``$TORCH_MODEL_ZOO`` environment variable. 30 | Args: 31 | url (string): URL of the object to download 32 | model_dir (string, optional): directory in which to save the object 33 | progress (bool, optional): whether or not to display a progress bar to stderr 34 | Example: 35 | >>> cached_file = maskrcnn_benchmark.utils.model_zoo.cache_url('https://s3.amazonaws.com/pytorch/models/resnet18-5c106cde.pth') 36 | """ 37 | if model_dir is None: 38 | torch_home = os.path.expanduser(os.getenv("TORCH_HOME", "~/.torch")) 39 | model_dir = os.getenv("TORCH_MODEL_ZOO", os.path.join(torch_home, "models")) 40 | if not os.path.exists(model_dir): 41 | os.makedirs(model_dir) 42 | parts = urlparse(url) 43 | filename = os.path.basename(parts.path) 44 | if filename == "model_final.pkl": 45 | # workaround as pre-trained Caffe2 models from Detectron have all the same filename 46 | # so make the full path the filename by replacing / with _ 47 | filename = parts.path.replace("/", "_") 48 | cached_file = os.path.join(model_dir, filename) 49 | if not os.path.exists(cached_file) and is_main_process(): 50 | sys.stderr.write('Downloading: "{}" to {}\n'.format(url, cached_file)) 51 | hash_prefix = HASH_REGEX.search(filename) 52 | if hash_prefix is not None: 53 | hash_prefix = hash_prefix.group(1) 54 | # workaround: Caffe2 models don't have a hash, but follow the R-50 convention, 55 | # which matches the hash PyTorch uses. So we skip the hash matching 56 | # if the hash_prefix is less than 6 characters 57 | if len(hash_prefix) < 6: 58 | hash_prefix = None 59 | _download_url_to_file(url, cached_file, hash_prefix, progress=progress) 60 | synchronize() 61 | return cached_file 62 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/mask_head/mask_head.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | from torch import nn 4 | 5 | from maskrcnn_benchmark.structures.bounding_box import BoxList 6 | 7 | from .roi_mask_feature_extractors import make_roi_mask_feature_extractor 8 | from .roi_mask_predictors import make_roi_mask_predictor 9 | from .inference import make_roi_mask_post_processor 10 | from .loss import make_roi_mask_loss_evaluator 11 | 12 | 13 | def keep_only_positive_boxes(boxes): 14 | """ 15 | Given a set of BoxList containing the `labels` field, 16 | return a set of BoxList for which `labels > 0`. 17 | 18 | Arguments: 19 | boxes (list of BoxList) 20 | """ 21 | assert isinstance(boxes, (list, tuple)) 22 | assert isinstance(boxes[0], BoxList) 23 | assert boxes[0].has_field("labels") 24 | positive_boxes = [] 25 | positive_inds = [] 26 | num_boxes = 0 27 | for boxes_per_image in boxes: 28 | labels = boxes_per_image.get_field("labels") 29 | inds_mask = labels > 0 30 | inds = inds_mask.nonzero().squeeze(1) 31 | positive_boxes.append(boxes_per_image[inds]) 32 | positive_inds.append(inds_mask) 33 | return positive_boxes, positive_inds 34 | 35 | 36 | class ROIMaskHead(torch.nn.Module): 37 | def __init__(self, cfg, in_channels): 38 | super(ROIMaskHead, self).__init__() 39 | self.cfg = cfg.clone() 40 | self.feature_extractor = make_roi_mask_feature_extractor(cfg, in_channels) 41 | self.predictor = make_roi_mask_predictor( 42 | cfg, self.feature_extractor.out_channels) 43 | self.post_processor = make_roi_mask_post_processor(cfg) 44 | self.loss_evaluator = make_roi_mask_loss_evaluator(cfg) 45 | 46 | def forward(self, features, proposals, targets=None): 47 | """ 48 | Arguments: 49 | features (list[Tensor]): feature-maps from possibly several levels 50 | proposals (list[BoxList]): proposal boxes 51 | targets (list[BoxList], optional): the ground-truth targets. 52 | 53 | Returns: 54 | x (Tensor): the result of the feature extractor 55 | proposals (list[BoxList]): during training, the original proposals 56 | are returned. During testing, the predicted boxlists are returned 57 | with the `mask` field set 58 | losses (dict[Tensor]): During training, returns the losses for the 59 | head. During testing, returns an empty dict. 60 | """ 61 | 62 | if self.training: 63 | # during training, only focus on positive boxes 64 | all_proposals = proposals 65 | proposals, positive_inds = keep_only_positive_boxes(proposals) 66 | if self.training and self.cfg.MODEL.ROI_MASK_HEAD.SHARE_BOX_FEATURE_EXTRACTOR: 67 | x = features 68 | x = x[torch.cat(positive_inds, dim=0)] 69 | else: 70 | x = self.feature_extractor(features, proposals) 71 | mask_logits = self.predictor(x) 72 | 73 | if not self.training: 74 | result = self.post_processor(mask_logits, proposals) 75 | return x, result, {} 76 | 77 | loss_mask = self.loss_evaluator(proposals, mask_logits, targets) 78 | 79 | return x, all_proposals, dict(loss_mask=loss_mask) 80 | 81 | 82 | def build_roi_mask_head(cfg, in_channels): 83 | return ROIMaskHead(cfg, in_channels) 84 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/roi_heads.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | 4 | from .box_head.box_head import build_roi_box_head 5 | from .mask_head.mask_head import build_roi_mask_head 6 | from .keypoint_head.keypoint_head import build_roi_keypoint_head 7 | 8 | 9 | class CombinedROIHeads(torch.nn.ModuleDict): 10 | """ 11 | Combines a set of individual heads (for box prediction or masks) into a single 12 | head. 13 | """ 14 | 15 | def __init__(self, cfg, heads): 16 | super(CombinedROIHeads, self).__init__(heads) 17 | self.cfg = cfg.clone() 18 | if cfg.MODEL.MASK_ON and cfg.MODEL.ROI_MASK_HEAD.SHARE_BOX_FEATURE_EXTRACTOR: 19 | self.mask.feature_extractor = self.box.feature_extractor 20 | if cfg.MODEL.KEYPOINT_ON and cfg.MODEL.ROI_KEYPOINT_HEAD.SHARE_BOX_FEATURE_EXTRACTOR: 21 | self.keypoint.feature_extractor = self.box.feature_extractor 22 | 23 | def forward(self, features, proposals, targets=None): 24 | losses = {} 25 | # TODO rename x to roi_box_features, if it doesn't increase memory consumption 26 | x, detections, loss_box = self.box(features, proposals, targets) 27 | losses.update(loss_box) 28 | if self.cfg.MODEL.MASK_ON: 29 | mask_features = features 30 | # optimization: during training, if we share the feature extractor between 31 | # the box and the mask heads, then we can reuse the features already computed 32 | if ( 33 | self.training 34 | and self.cfg.MODEL.ROI_MASK_HEAD.SHARE_BOX_FEATURE_EXTRACTOR 35 | ): 36 | mask_features = x 37 | # During training, self.box() will return the unaltered proposals as "detections" 38 | # this makes the API consistent during training and testing 39 | x, detections, loss_mask = self.mask(mask_features, detections, targets) 40 | losses.update(loss_mask) 41 | 42 | if self.cfg.MODEL.KEYPOINT_ON: 43 | keypoint_features = features 44 | # optimization: during training, if we share the feature extractor between 45 | # the box and the mask heads, then we can reuse the features already computed 46 | if ( 47 | self.training 48 | and self.cfg.MODEL.ROI_KEYPOINT_HEAD.SHARE_BOX_FEATURE_EXTRACTOR 49 | ): 50 | keypoint_features = x 51 | # During training, self.box() will return the unaltered proposals as "detections" 52 | # this makes the API consistent during training and testing 53 | x, detections, loss_keypoint = self.keypoint(keypoint_features, detections, targets) 54 | losses.update(loss_keypoint) 55 | return x, detections, losses 56 | 57 | 58 | def build_roi_heads(cfg, in_channels): 59 | # individually create the heads, that will be combined together 60 | # afterwards 61 | roi_heads = [] 62 | if cfg.MODEL.RETINANET_ON: 63 | return [] 64 | 65 | if not cfg.MODEL.RPN_ONLY: 66 | roi_heads.append(("box", build_roi_box_head(cfg, in_channels))) 67 | if cfg.MODEL.MASK_ON: 68 | roi_heads.append(("mask", build_roi_mask_head(cfg, in_channels))) 69 | if cfg.MODEL.KEYPOINT_ON: 70 | roi_heads.append(("keypoint", build_roi_keypoint_head(cfg, in_channels))) 71 | 72 | # combine individual heads in a single module 73 | if roi_heads: 74 | roi_heads = CombinedROIHeads(cfg, roi_heads) 75 | 76 | return roi_heads 77 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/box_coder.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import math 3 | 4 | import torch 5 | 6 | 7 | class BoxCoder(object): 8 | """ 9 | This class encodes and decodes a set of bounding boxes into 10 | the representation used for training the regressors. 11 | """ 12 | 13 | def __init__(self, weights, bbox_xform_clip=math.log(1000. / 16)): 14 | """ 15 | Arguments: 16 | weights (4-element tuple) 17 | bbox_xform_clip (float) 18 | """ 19 | self.weights = weights 20 | self.bbox_xform_clip = bbox_xform_clip 21 | 22 | def encode(self, reference_boxes, proposals): 23 | """ 24 | Encode a set of proposals with respect to some 25 | reference boxes 26 | 27 | Arguments: 28 | reference_boxes (Tensor): reference boxes 29 | proposals (Tensor): boxes to be encoded 30 | """ 31 | 32 | TO_REMOVE = 1 # TODO remove 33 | ex_widths = proposals[:, 2] - proposals[:, 0] + TO_REMOVE 34 | ex_heights = proposals[:, 3] - proposals[:, 1] + TO_REMOVE 35 | ex_ctr_x = proposals[:, 0] + 0.5 * ex_widths 36 | ex_ctr_y = proposals[:, 1] + 0.5 * ex_heights 37 | 38 | gt_widths = reference_boxes[:, 2] - reference_boxes[:, 0] + TO_REMOVE 39 | gt_heights = reference_boxes[:, 3] - reference_boxes[:, 1] + TO_REMOVE 40 | gt_ctr_x = reference_boxes[:, 0] + 0.5 * gt_widths 41 | gt_ctr_y = reference_boxes[:, 1] + 0.5 * gt_heights 42 | 43 | wx, wy, ww, wh = self.weights 44 | targets_dx = wx * (gt_ctr_x - ex_ctr_x) / ex_widths 45 | targets_dy = wy * (gt_ctr_y - ex_ctr_y) / ex_heights 46 | targets_dw = ww * torch.log(gt_widths / ex_widths) 47 | targets_dh = wh * torch.log(gt_heights / ex_heights) 48 | 49 | targets = torch.stack((targets_dx, targets_dy, targets_dw, targets_dh), dim=1) 50 | return targets 51 | 52 | def decode(self, rel_codes, boxes): 53 | """ 54 | From a set of original boxes and encoded relative box offsets, 55 | get the decoded boxes. 56 | 57 | Arguments: 58 | rel_codes (Tensor): encoded boxes 59 | boxes (Tensor): reference boxes. 60 | """ 61 | 62 | boxes = boxes.to(rel_codes.dtype) 63 | 64 | TO_REMOVE = 1 # TODO remove 65 | widths = boxes[:, 2] - boxes[:, 0] + TO_REMOVE 66 | heights = boxes[:, 3] - boxes[:, 1] + TO_REMOVE 67 | ctr_x = boxes[:, 0] + 0.5 * widths 68 | ctr_y = boxes[:, 1] + 0.5 * heights 69 | 70 | wx, wy, ww, wh = self.weights 71 | dx = rel_codes[:, 0::4] / wx 72 | dy = rel_codes[:, 1::4] / wy 73 | dw = rel_codes[:, 2::4] / ww 74 | dh = rel_codes[:, 3::4] / wh 75 | 76 | # Prevent sending too large values into torch.exp() 77 | dw = torch.clamp(dw, max=self.bbox_xform_clip) 78 | dh = torch.clamp(dh, max=self.bbox_xform_clip) 79 | 80 | pred_ctr_x = dx * widths[:, None] + ctr_x[:, None] 81 | pred_ctr_y = dy * heights[:, None] + ctr_y[:, None] 82 | pred_w = torch.exp(dw) * widths[:, None] 83 | pred_h = torch.exp(dh) * heights[:, None] 84 | 85 | pred_boxes = torch.zeros_like(rel_codes) 86 | # x1 87 | pred_boxes[:, 0::4] = pred_ctr_x - 0.5 * pred_w 88 | # y1 89 | pred_boxes[:, 1::4] = pred_ctr_y - 0.5 * pred_h 90 | # x2 (note: "- 1" is correct; don't be fooled by the asymmetry) 91 | pred_boxes[:, 2::4] = pred_ctr_x + 0.5 * pred_w - 1 92 | # y2 (note: "- 1" is correct; don't be fooled by the asymmetry) 93 | pred_boxes[:, 3::4] = pred_ctr_y + 0.5 * pred_h - 1 94 | 95 | return pred_boxes 96 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/model_serialization.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | from collections import OrderedDict 3 | import logging 4 | 5 | import torch 6 | 7 | from maskrcnn_benchmark.utils.imports import import_file 8 | 9 | 10 | def align_and_update_state_dicts(model_state_dict, loaded_state_dict): 11 | """ 12 | Strategy: suppose that the models that we will create will have prefixes appended 13 | to each of its keys, for example due to an extra level of nesting that the original 14 | pre-trained weights from ImageNet won't contain. For example, model.state_dict() 15 | might return backbone[0].body.res2.conv1.weight, while the pre-trained model contains 16 | res2.conv1.weight. We thus want to match both parameters together. 17 | For that, we look for each model weight, look among all loaded keys if there is one 18 | that is a suffix of the current weight name, and use it if that's the case. 19 | If multiple matches exist, take the one with longest size 20 | of the corresponding name. For example, for the same model as before, the pretrained 21 | weight file can contain both res2.conv1.weight, as well as conv1.weight. In this case, 22 | we want to match backbone[0].body.conv1.weight to conv1.weight, and 23 | backbone[0].body.res2.conv1.weight to res2.conv1.weight. 24 | """ 25 | current_keys = sorted(list(model_state_dict.keys())) 26 | loaded_keys = sorted(list(loaded_state_dict.keys())) 27 | # get a matrix of string matches, where each (i, j) entry correspond to the size of the 28 | # loaded_key string, if it matches 29 | match_matrix = [ 30 | len(j) if i.endswith(j) else 0 for i in current_keys for j in loaded_keys 31 | ] 32 | match_matrix = torch.as_tensor(match_matrix).view( 33 | len(current_keys), len(loaded_keys) 34 | ) 35 | max_match_size, idxs = match_matrix.max(1) 36 | # remove indices that correspond to no-match 37 | idxs[max_match_size == 0] = -1 38 | 39 | # used for logging 40 | max_size = max([len(key) for key in current_keys]) if current_keys else 1 41 | max_size_loaded = max([len(key) for key in loaded_keys]) if loaded_keys else 1 42 | log_str_template = "{: <{}} loaded from {: <{}} of shape {}" 43 | logger = logging.getLogger(__name__) 44 | for idx_new, idx_old in enumerate(idxs.tolist()): 45 | if idx_old == -1: 46 | continue 47 | key = current_keys[idx_new] 48 | key_old = loaded_keys[idx_old] 49 | model_state_dict[key] = loaded_state_dict[key_old] 50 | logger.info( 51 | log_str_template.format( 52 | key, 53 | max_size, 54 | key_old, 55 | max_size_loaded, 56 | tuple(loaded_state_dict[key_old].shape), 57 | ) 58 | ) 59 | 60 | 61 | def strip_prefix_if_present(state_dict, prefix): 62 | keys = sorted(state_dict.keys()) 63 | if not all(key.startswith(prefix) for key in keys): 64 | return state_dict 65 | stripped_state_dict = OrderedDict() 66 | for key, value in state_dict.items(): 67 | stripped_state_dict[key.replace(prefix, "")] = value 68 | return stripped_state_dict 69 | 70 | 71 | def load_state_dict(model, loaded_state_dict): 72 | model_state_dict = model.state_dict() 73 | # if the state_dict comes from a model that was wrapped in a 74 | # DataParallel or DistributedDataParallel during serialization, 75 | # remove the "module" prefix before performing the matching 76 | loaded_state_dict = strip_prefix_if_present(loaded_state_dict, prefix="module.") 77 | align_and_update_state_dicts(model_state_dict, loaded_state_dict) 78 | 79 | # use strict loading 80 | model.load_state_dict(model_state_dict) 81 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/utils/comm.py: -------------------------------------------------------------------------------- 1 | """ 2 | This file contains primitives for multi-gpu communication. 3 | This is useful when doing distributed training. 4 | """ 5 | 6 | import pickle 7 | import time 8 | 9 | import torch 10 | import torch.distributed as dist 11 | 12 | 13 | def get_world_size(): 14 | if not dist.is_available(): 15 | return 1 16 | if not dist.is_initialized(): 17 | return 1 18 | return dist.get_world_size() 19 | 20 | 21 | def get_rank(): 22 | if not dist.is_available(): 23 | return 0 24 | if not dist.is_initialized(): 25 | return 0 26 | return dist.get_rank() 27 | 28 | 29 | def is_main_process(): 30 | return get_rank() == 0 31 | 32 | 33 | def synchronize(): 34 | """ 35 | Helper function to synchronize (barrier) among all processes when 36 | using distributed training 37 | """ 38 | if not dist.is_available(): 39 | return 40 | if not dist.is_initialized(): 41 | return 42 | world_size = dist.get_world_size() 43 | if world_size == 1: 44 | return 45 | dist.barrier() 46 | 47 | 48 | def all_gather(data): 49 | """ 50 | Run all_gather on arbitrary picklable data (not necessarily tensors) 51 | Args: 52 | data: any picklable object 53 | Returns: 54 | list[data]: list of data gathered from each rank 55 | """ 56 | world_size = get_world_size() 57 | if world_size == 1: 58 | return [data] 59 | 60 | # serialized to a Tensor 61 | buffer = pickle.dumps(data) 62 | storage = torch.ByteStorage.from_buffer(buffer) 63 | tensor = torch.ByteTensor(storage).to("cuda") 64 | 65 | # obtain Tensor size of each rank 66 | local_size = torch.LongTensor([tensor.numel()]).to("cuda") 67 | size_list = [torch.LongTensor([0]).to("cuda") for _ in range(world_size)] 68 | dist.all_gather(size_list, local_size) 69 | size_list = [int(size.item()) for size in size_list] 70 | max_size = max(size_list) 71 | 72 | # receiving Tensor from all ranks 73 | # we pad the tensor because torch all_gather does not support 74 | # gathering tensors of different shapes 75 | tensor_list = [] 76 | for _ in size_list: 77 | tensor_list.append(torch.ByteTensor(size=(max_size,)).to("cuda")) 78 | if local_size != max_size: 79 | padding = torch.ByteTensor(size=(max_size - local_size,)).to("cuda") 80 | tensor = torch.cat((tensor, padding), dim=0) 81 | dist.all_gather(tensor_list, tensor) 82 | 83 | data_list = [] 84 | for size, tensor in zip(size_list, tensor_list): 85 | buffer = tensor.cpu().numpy().tobytes()[:size] 86 | data_list.append(pickle.loads(buffer)) 87 | 88 | return data_list 89 | 90 | 91 | def reduce_dict(input_dict, average=True): 92 | """ 93 | Args: 94 | input_dict (dict): all the values will be reduced 95 | average (bool): whether to do average or sum 96 | Reduce the values in the dictionary from all processes so that process with rank 97 | 0 has the averaged results. Returns a dict with the same fields as 98 | input_dict, after reduction. 99 | """ 100 | world_size = get_world_size() 101 | if world_size < 2: 102 | return input_dict 103 | with torch.no_grad(): 104 | names = [] 105 | values = [] 106 | # sort the keys so that they are consistent across processes 107 | for k in sorted(input_dict.keys()): 108 | names.append(k) 109 | values.append(input_dict[k]) 110 | values = torch.stack(values, dim=0) 111 | dist.reduce(values, dst=0) 112 | if dist.get_rank() == 0 and average: 113 | # only main process gets accumulated, so only divide by 114 | # world_size in this case 115 | values /= world_size 116 | reduced_dict = {k: v for k, v in zip(names, values)} 117 | return reduced_dict 118 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/rpn/retinanet/loss.py: -------------------------------------------------------------------------------- 1 | """ 2 | This file contains specific functions for computing losses on the RetinaNet 3 | file 4 | """ 5 | 6 | import torch 7 | from torch.nn import functional as F 8 | 9 | from ..utils import concat_box_prediction_layers 10 | 11 | from maskrcnn_benchmark.layers import smooth_l1_loss 12 | from maskrcnn_benchmark.layers import SigmoidFocalLoss 13 | from maskrcnn_benchmark.modeling.matcher import Matcher 14 | from maskrcnn_benchmark.modeling.utils import cat 15 | from maskrcnn_benchmark.structures.boxlist_ops import boxlist_iou 16 | from maskrcnn_benchmark.structures.boxlist_ops import cat_boxlist 17 | from maskrcnn_benchmark.modeling.rpn.loss import RPNLossComputation 18 | 19 | class RetinaNetLossComputation(RPNLossComputation): 20 | """ 21 | This class computes the RetinaNet loss. 22 | """ 23 | 24 | def __init__(self, proposal_matcher, box_coder, 25 | generate_labels_func, 26 | sigmoid_focal_loss, 27 | bbox_reg_beta=0.11, 28 | regress_norm=1.0): 29 | """ 30 | Arguments: 31 | proposal_matcher (Matcher) 32 | box_coder (BoxCoder) 33 | """ 34 | self.proposal_matcher = proposal_matcher 35 | self.box_coder = box_coder 36 | self.box_cls_loss_func = sigmoid_focal_loss 37 | self.bbox_reg_beta = bbox_reg_beta 38 | self.copied_fields = ['labels'] 39 | self.generate_labels_func = generate_labels_func 40 | self.discard_cases = ['between_thresholds'] 41 | self.regress_norm = regress_norm 42 | 43 | def __call__(self, anchors, box_cls, box_regression, targets): 44 | """ 45 | Arguments: 46 | anchors (list[BoxList]) 47 | box_cls (list[Tensor]) 48 | box_regression (list[Tensor]) 49 | targets (list[BoxList]) 50 | 51 | Returns: 52 | retinanet_cls_loss (Tensor) 53 | retinanet_regression_loss (Tensor 54 | """ 55 | anchors = [cat_boxlist(anchors_per_image) for anchors_per_image in anchors] 56 | labels, regression_targets = self.prepare_targets(anchors, targets) 57 | 58 | N = len(labels) 59 | box_cls, box_regression = \ 60 | concat_box_prediction_layers(box_cls, box_regression) 61 | 62 | labels = torch.cat(labels, dim=0) 63 | regression_targets = torch.cat(regression_targets, dim=0) 64 | pos_inds = torch.nonzero(labels > 0).squeeze(1) 65 | 66 | retinanet_regression_loss = smooth_l1_loss( 67 | box_regression[pos_inds], 68 | regression_targets[pos_inds], 69 | beta=self.bbox_reg_beta, 70 | size_average=False, 71 | ) / (max(1, pos_inds.numel() * self.regress_norm)) 72 | 73 | labels = labels.int() 74 | 75 | retinanet_cls_loss = self.box_cls_loss_func( 76 | box_cls, 77 | labels 78 | ) / (pos_inds.numel() + N) 79 | 80 | return retinanet_cls_loss, retinanet_regression_loss 81 | 82 | 83 | def generate_retinanet_labels(matched_targets): 84 | labels_per_image = matched_targets.get_field("labels") 85 | return labels_per_image 86 | 87 | 88 | def make_retinanet_loss_evaluator(cfg, box_coder): 89 | matcher = Matcher( 90 | cfg.MODEL.RETINANET.FG_IOU_THRESHOLD, 91 | cfg.MODEL.RETINANET.BG_IOU_THRESHOLD, 92 | allow_low_quality_matches=True, 93 | ) 94 | sigmoid_focal_loss = SigmoidFocalLoss( 95 | cfg.MODEL.RETINANET.LOSS_GAMMA, 96 | cfg.MODEL.RETINANET.LOSS_ALPHA 97 | ) 98 | 99 | loss_evaluator = RetinaNetLossComputation( 100 | matcher, 101 | box_coder, 102 | generate_retinanet_labels, 103 | sigmoid_focal_loss, 104 | bbox_reg_beta = cfg.MODEL.RETINANET.BBOX_REG_BETA, 105 | regress_norm = cfg.MODEL.RETINANET.BBOX_REG_WEIGHT, 106 | ) 107 | return loss_evaluator 108 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/cuda/deform_pool_cuda.cu: -------------------------------------------------------------------------------- 1 | // modify from 2 | // https://github.com/chengdazhi/Deformable-Convolution-V2-PyTorch/blob/mmdetection/mmdet/ops/dcn/src/modulated_dcn_cuda.c 3 | 4 | // based on 5 | // author: Charles Shang 6 | // https://github.com/torch/cunn/blob/master/lib/THCUNN/generic/SpatialConvolutionMM.cu 7 | 8 | #include 9 | #include 10 | 11 | #include 12 | #include 13 | 14 | #include 15 | #include 16 | #include 17 | 18 | 19 | void DeformablePSROIPoolForward( 20 | const at::Tensor data, const at::Tensor bbox, const at::Tensor trans, 21 | at::Tensor out, at::Tensor top_count, const int batch, const int channels, 22 | const int height, const int width, const int num_bbox, 23 | const int channels_trans, const int no_trans, const float spatial_scale, 24 | const int output_dim, const int group_size, const int pooled_size, 25 | const int part_size, const int sample_per_part, const float trans_std); 26 | 27 | void DeformablePSROIPoolBackwardAcc( 28 | const at::Tensor out_grad, const at::Tensor data, const at::Tensor bbox, 29 | const at::Tensor trans, const at::Tensor top_count, at::Tensor in_grad, 30 | at::Tensor trans_grad, const int batch, const int channels, 31 | const int height, const int width, const int num_bbox, 32 | const int channels_trans, const int no_trans, const float spatial_scale, 33 | const int output_dim, const int group_size, const int pooled_size, 34 | const int part_size, const int sample_per_part, const float trans_std); 35 | 36 | void deform_psroi_pooling_cuda_forward( 37 | at::Tensor input, at::Tensor bbox, at::Tensor trans, at::Tensor out, 38 | at::Tensor top_count, const int no_trans, const float spatial_scale, 39 | const int output_dim, const int group_size, const int pooled_size, 40 | const int part_size, const int sample_per_part, const float trans_std) 41 | { 42 | AT_CHECK(input.is_contiguous(), "input tensor has to be contiguous"); 43 | 44 | const int batch = input.size(0); 45 | const int channels = input.size(1); 46 | const int height = input.size(2); 47 | const int width = input.size(3); 48 | const int channels_trans = no_trans ? 2 : trans.size(1); 49 | 50 | const int num_bbox = bbox.size(0); 51 | if (num_bbox != out.size(0)) 52 | AT_ERROR("Output shape and bbox number wont match: (%d vs %d).", 53 | out.size(0), num_bbox); 54 | 55 | DeformablePSROIPoolForward( 56 | input, bbox, trans, out, top_count, batch, channels, height, width, 57 | num_bbox, channels_trans, no_trans, spatial_scale, output_dim, group_size, 58 | pooled_size, part_size, sample_per_part, trans_std); 59 | } 60 | 61 | void deform_psroi_pooling_cuda_backward( 62 | at::Tensor out_grad, at::Tensor input, at::Tensor bbox, at::Tensor trans, 63 | at::Tensor top_count, at::Tensor input_grad, at::Tensor trans_grad, 64 | const int no_trans, const float spatial_scale, const int output_dim, 65 | const int group_size, const int pooled_size, const int part_size, 66 | const int sample_per_part, const float trans_std) 67 | { 68 | AT_CHECK(out_grad.is_contiguous(), "out_grad tensor has to be contiguous"); 69 | AT_CHECK(input.is_contiguous(), "input tensor has to be contiguous"); 70 | 71 | const int batch = input.size(0); 72 | const int channels = input.size(1); 73 | const int height = input.size(2); 74 | const int width = input.size(3); 75 | const int channels_trans = no_trans ? 2 : trans.size(1); 76 | 77 | const int num_bbox = bbox.size(0); 78 | if (num_bbox != out_grad.size(0)) 79 | AT_ERROR("Output shape and bbox number wont match: (%d vs %d).", 80 | out_grad.size(0), num_bbox); 81 | 82 | DeformablePSROIPoolBackwardAcc( 83 | out_grad, input, bbox, trans, top_count, input_grad, trans_grad, batch, 84 | channels, height, width, num_bbox, channels_trans, no_trans, 85 | spatial_scale, output_dim, group_size, pooled_size, part_size, 86 | sample_per_part, trans_std); 87 | } 88 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/transforms/transforms.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import random 3 | 4 | import torch 5 | import torchvision 6 | from torchvision.transforms import functional as F 7 | 8 | 9 | class Compose(object): 10 | def __init__(self, transforms): 11 | self.transforms = transforms 12 | 13 | def __call__(self, image, target): 14 | for t in self.transforms: 15 | image, target = t(image, target) 16 | return image, target 17 | 18 | def __repr__(self): 19 | format_string = self.__class__.__name__ + "(" 20 | for t in self.transforms: 21 | format_string += "\n" 22 | format_string += " {0}".format(t) 23 | format_string += "\n)" 24 | return format_string 25 | 26 | 27 | class Resize(object): 28 | def __init__(self, min_size, max_size): 29 | if not isinstance(min_size, (list, tuple)): 30 | min_size = (min_size,) 31 | self.min_size = min_size 32 | self.max_size = max_size 33 | 34 | # modified from torchvision to add support for max size 35 | def get_size(self, image_size): 36 | w, h = image_size 37 | size = random.choice(self.min_size) 38 | max_size = self.max_size 39 | if max_size is not None: 40 | min_original_size = float(min((w, h))) 41 | max_original_size = float(max((w, h))) 42 | if max_original_size / min_original_size * size > max_size: 43 | size = int(round(max_size * min_original_size / max_original_size)) 44 | 45 | if (w <= h and w == size) or (h <= w and h == size): 46 | return (h, w) 47 | 48 | if w < h: 49 | ow = size 50 | oh = int(size * h / w) 51 | else: 52 | oh = size 53 | ow = int(size * w / h) 54 | 55 | return (oh, ow) 56 | 57 | def __call__(self, image, target=None): 58 | size = self.get_size(image.size) 59 | image = F.resize(image, size) 60 | if target is None: 61 | return image 62 | target = target.resize(image.size) 63 | return image, target 64 | 65 | 66 | class RandomHorizontalFlip(object): 67 | def __init__(self, prob=0.5): 68 | self.prob = prob 69 | 70 | def __call__(self, image, target): 71 | if random.random() < self.prob: 72 | image = F.hflip(image) 73 | target = target.transpose(0) 74 | return image, target 75 | 76 | class RandomVerticalFlip(object): 77 | def __init__(self, prob=0.5): 78 | self.prob = prob 79 | 80 | def __call__(self, image, target): 81 | if random.random() < self.prob: 82 | image = F.vflip(image) 83 | target = target.transpose(1) 84 | return image, target 85 | 86 | class ColorJitter(object): 87 | def __init__(self, 88 | brightness=None, 89 | contrast=None, 90 | saturation=None, 91 | hue=None, 92 | ): 93 | self.color_jitter = torchvision.transforms.ColorJitter( 94 | brightness=brightness, 95 | contrast=contrast, 96 | saturation=saturation, 97 | hue=hue,) 98 | 99 | def __call__(self, image, target): 100 | image = self.color_jitter(image) 101 | return image, target 102 | 103 | 104 | class ToTensor(object): 105 | def __call__(self, image, target): 106 | return F.to_tensor(image), target 107 | 108 | 109 | class Normalize(object): 110 | def __init__(self, mean, std, to_bgr255=True): 111 | self.mean = mean 112 | self.std = std 113 | self.to_bgr255 = to_bgr255 114 | 115 | def __call__(self, image, target=None): 116 | if self.to_bgr255: 117 | image = image[[2, 1, 0]] * 255 118 | image = F.normalize(image, mean=self.mean, std=self.std) 119 | if target is None: 120 | return image 121 | return image, target 122 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/make_layers.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | """ 3 | Miscellaneous utility functions 4 | """ 5 | 6 | import torch 7 | from torch import nn 8 | from torch.nn import functional as F 9 | from maskrcnn_benchmark.config import cfg 10 | from maskrcnn_benchmark.layers import Conv2d 11 | from maskrcnn_benchmark.modeling.poolers import Pooler 12 | 13 | 14 | def get_group_gn(dim, dim_per_gp, num_groups): 15 | """get number of groups used by GroupNorm, based on number of channels.""" 16 | assert dim_per_gp == -1 or num_groups == -1, \ 17 | "GroupNorm: can only specify G or C/G." 18 | 19 | if dim_per_gp > 0: 20 | assert dim % dim_per_gp == 0, \ 21 | "dim: {}, dim_per_gp: {}".format(dim, dim_per_gp) 22 | group_gn = dim // dim_per_gp 23 | else: 24 | assert dim % num_groups == 0, \ 25 | "dim: {}, num_groups: {}".format(dim, num_groups) 26 | group_gn = num_groups 27 | 28 | return group_gn 29 | 30 | 31 | def group_norm(out_channels, affine=True, divisor=1): 32 | out_channels = out_channels // divisor 33 | dim_per_gp = cfg.MODEL.GROUP_NORM.DIM_PER_GP // divisor 34 | num_groups = cfg.MODEL.GROUP_NORM.NUM_GROUPS // divisor 35 | eps = cfg.MODEL.GROUP_NORM.EPSILON # default: 1e-5 36 | return torch.nn.GroupNorm( 37 | get_group_gn(out_channels, dim_per_gp, num_groups), 38 | out_channels, 39 | eps, 40 | affine 41 | ) 42 | 43 | 44 | def make_conv3x3( 45 | in_channels, 46 | out_channels, 47 | dilation=1, 48 | stride=1, 49 | use_gn=False, 50 | use_relu=False, 51 | kaiming_init=True 52 | ): 53 | conv = Conv2d( 54 | in_channels, 55 | out_channels, 56 | kernel_size=3, 57 | stride=stride, 58 | padding=dilation, 59 | dilation=dilation, 60 | bias=False if use_gn else True 61 | ) 62 | if kaiming_init: 63 | nn.init.kaiming_normal_( 64 | conv.weight, mode="fan_out", nonlinearity="relu" 65 | ) 66 | else: 67 | torch.nn.init.normal_(conv.weight, std=0.01) 68 | if not use_gn: 69 | nn.init.constant_(conv.bias, 0) 70 | module = [conv,] 71 | if use_gn: 72 | module.append(group_norm(out_channels)) 73 | if use_relu: 74 | module.append(nn.ReLU(inplace=True)) 75 | if len(module) > 1: 76 | return nn.Sequential(*module) 77 | return conv 78 | 79 | 80 | def make_fc(dim_in, hidden_dim, use_gn=False): 81 | ''' 82 | Caffe2 implementation uses XavierFill, which in fact 83 | corresponds to kaiming_uniform_ in PyTorch 84 | ''' 85 | if use_gn: 86 | fc = nn.Linear(dim_in, hidden_dim, bias=False) 87 | nn.init.kaiming_uniform_(fc.weight, a=1) 88 | return nn.Sequential(fc, group_norm(hidden_dim)) 89 | fc = nn.Linear(dim_in, hidden_dim) 90 | nn.init.kaiming_uniform_(fc.weight, a=1) 91 | nn.init.constant_(fc.bias, 0) 92 | return fc 93 | 94 | 95 | def conv_with_kaiming_uniform(use_gn=False, use_relu=False): 96 | def make_conv( 97 | in_channels, out_channels, kernel_size, stride=1, dilation=1 98 | ): 99 | conv = Conv2d( 100 | in_channels, 101 | out_channels, 102 | kernel_size=kernel_size, 103 | stride=stride, 104 | padding=dilation * (kernel_size - 1) // 2, 105 | dilation=dilation, 106 | bias=False if use_gn else True 107 | ) 108 | # Caffe2 implementation uses XavierFill, which in fact 109 | # corresponds to kaiming_uniform_ in PyTorch 110 | nn.init.kaiming_uniform_(conv.weight, a=1) 111 | if not use_gn: 112 | nn.init.constant_(conv.bias, 0) 113 | module = [conv,] 114 | if use_gn: 115 | module.append(group_norm(out_channels)) 116 | if use_relu: 117 | module.append(nn.ReLU(inplace=True)) 118 | if len(module) > 1: 119 | return nn.Sequential(*module) 120 | return conv 121 | 122 | return make_conv 123 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/coco.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | import torchvision 4 | 5 | from maskrcnn_benchmark.structures.bounding_box import BoxList 6 | from maskrcnn_benchmark.structures.segmentation_mask import SegmentationMask 7 | from maskrcnn_benchmark.structures.keypoint import PersonKeypoints 8 | 9 | 10 | min_keypoints_per_image = 10 11 | 12 | 13 | def _count_visible_keypoints(anno): 14 | return sum(sum(1 for v in ann["keypoints"][2::3] if v > 0) for ann in anno) 15 | 16 | 17 | def _has_only_empty_bbox(anno): 18 | return all(any(o <= 1 for o in obj["bbox"][2:]) for obj in anno) 19 | 20 | 21 | def has_valid_annotation(anno): 22 | # if it's empty, there is no annotation 23 | if len(anno) == 0: 24 | return False 25 | # if all boxes have close to zero area, there is no annotation 26 | if _has_only_empty_bbox(anno): 27 | return False 28 | # keypoints task have a slight different critera for considering 29 | # if an annotation is valid 30 | if "keypoints" not in anno[0]: 31 | return True 32 | # for keypoint detection tasks, only consider valid images those 33 | # containing at least min_keypoints_per_image 34 | if _count_visible_keypoints(anno) >= min_keypoints_per_image: 35 | return True 36 | return False 37 | 38 | 39 | class COCODataset(torchvision.datasets.coco.CocoDetection): 40 | def __init__( 41 | self, ann_file, root, remove_images_without_annotations, transforms=None 42 | ): 43 | super(COCODataset, self).__init__(root, ann_file) 44 | # sort indices for reproducible results 45 | self.ids = sorted(self.ids) 46 | 47 | # filter images without detection annotations 48 | if remove_images_without_annotations: 49 | ids = [] 50 | for img_id in self.ids: 51 | ann_ids = self.coco.getAnnIds(imgIds=img_id, iscrowd=None) 52 | anno = self.coco.loadAnns(ann_ids) 53 | if has_valid_annotation(anno): 54 | ids.append(img_id) 55 | self.ids = ids 56 | 57 | self.categories = {cat['id']: cat['name'] for cat in self.coco.cats.values()} 58 | 59 | self.json_category_id_to_contiguous_id = { 60 | v: i + 1 for i, v in enumerate(self.coco.getCatIds()) 61 | } 62 | self.contiguous_category_id_to_json_id = { 63 | v: k for k, v in self.json_category_id_to_contiguous_id.items() 64 | } 65 | self.id_to_img_map = {k: v for k, v in enumerate(self.ids)} 66 | self._transforms = transforms 67 | 68 | def __getitem__(self, idx): 69 | img, anno = super(COCODataset, self).__getitem__(idx) 70 | 71 | # filter crowd annotations 72 | # TODO might be better to add an extra field 73 | anno = [obj for obj in anno if obj["iscrowd"] == 0] 74 | 75 | boxes = [obj["bbox"] for obj in anno] 76 | boxes = torch.as_tensor(boxes).reshape(-1, 4) # guard against no boxes 77 | target = BoxList(boxes, img.size, mode="xywh").convert("xyxy") 78 | 79 | classes = [obj["category_id"] for obj in anno] 80 | classes = [self.json_category_id_to_contiguous_id[c] for c in classes] 81 | classes = torch.tensor(classes) 82 | target.add_field("labels", classes) 83 | 84 | if anno and "segmentation" in anno[0]: 85 | masks = [obj["segmentation"] for obj in anno] 86 | masks = SegmentationMask(masks, img.size, mode='poly') 87 | target.add_field("masks", masks) 88 | 89 | if anno and "keypoints" in anno[0]: 90 | keypoints = [obj["keypoints"] for obj in anno] 91 | keypoints = PersonKeypoints(keypoints, img.size) 92 | target.add_field("keypoints", keypoints) 93 | 94 | target = target.clip_to_image(remove_empty=True) 95 | 96 | if self._transforms is not None: 97 | img, target = self._transforms(img, target) 98 | 99 | return img, target, idx 100 | 101 | def get_img_info(self, index): 102 | img_id = self.id_to_img_map[index] 103 | img_data = self.coco.imgs[img_id] 104 | return img_data 105 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/structures/boxlist_ops.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | 4 | from .bounding_box import BoxList 5 | 6 | from maskrcnn_benchmark.layers import nms as _box_nms 7 | 8 | 9 | def boxlist_nms(boxlist, nms_thresh, max_proposals=-1, score_field="scores"): 10 | """ 11 | Performs non-maximum suppression on a boxlist, with scores specified 12 | in a boxlist field via score_field. 13 | 14 | Arguments: 15 | boxlist(BoxList) 16 | nms_thresh (float) 17 | max_proposals (int): if > 0, then only the top max_proposals are kept 18 | after non-maximum suppression 19 | score_field (str) 20 | """ 21 | if nms_thresh <= 0: 22 | return boxlist 23 | mode = boxlist.mode 24 | boxlist = boxlist.convert("xyxy") 25 | boxes = boxlist.bbox 26 | score = boxlist.get_field(score_field) 27 | keep = _box_nms(boxes, score, nms_thresh) 28 | if max_proposals > 0: 29 | keep = keep[: max_proposals] 30 | boxlist = boxlist[keep] 31 | return boxlist.convert(mode) 32 | 33 | 34 | def remove_small_boxes(boxlist, min_size): 35 | """ 36 | Only keep boxes with both sides >= min_size 37 | 38 | Arguments: 39 | boxlist (Boxlist) 40 | min_size (int) 41 | """ 42 | # TODO maybe add an API for querying the ws / hs 43 | xywh_boxes = boxlist.convert("xywh").bbox 44 | _, _, ws, hs = xywh_boxes.unbind(dim=1) 45 | keep = ( 46 | (ws >= min_size) & (hs >= min_size) 47 | ).nonzero().squeeze(1) 48 | return boxlist[keep] 49 | 50 | 51 | # implementation from https://github.com/kuangliu/torchcv/blob/master/torchcv/utils/box.py 52 | # with slight modifications 53 | def boxlist_iou(boxlist1, boxlist2): 54 | """Compute the intersection over union of two set of boxes. 55 | The box order must be (xmin, ymin, xmax, ymax). 56 | 57 | Arguments: 58 | box1: (BoxList) bounding boxes, sized [N,4]. 59 | box2: (BoxList) bounding boxes, sized [M,4]. 60 | 61 | Returns: 62 | (tensor) iou, sized [N,M]. 63 | 64 | Reference: 65 | https://github.com/chainer/chainercv/blob/master/chainercv/utils/bbox/bbox_iou.py 66 | """ 67 | if boxlist1.size != boxlist2.size: 68 | raise RuntimeError( 69 | "boxlists should have same image size, got {}, {}".format(boxlist1, boxlist2)) 70 | boxlist1 = boxlist1.convert("xyxy") 71 | boxlist2 = boxlist2.convert("xyxy") 72 | N = len(boxlist1) 73 | M = len(boxlist2) 74 | 75 | area1 = boxlist1.area() 76 | area2 = boxlist2.area() 77 | 78 | box1, box2 = boxlist1.bbox, boxlist2.bbox 79 | 80 | lt = torch.max(box1[:, None, :2], box2[:, :2]) # [N,M,2] 81 | rb = torch.min(box1[:, None, 2:], box2[:, 2:]) # [N,M,2] 82 | 83 | TO_REMOVE = 1 84 | 85 | wh = (rb - lt + TO_REMOVE).clamp(min=0) # [N,M,2] 86 | inter = wh[:, :, 0] * wh[:, :, 1] # [N,M] 87 | 88 | iou = inter / (area1[:, None] + area2 - inter) 89 | return iou 90 | 91 | 92 | # TODO redundant, remove 93 | def _cat(tensors, dim=0): 94 | """ 95 | Efficient version of torch.cat that avoids a copy if there is only a single element in a list 96 | """ 97 | assert isinstance(tensors, (list, tuple)) 98 | if len(tensors) == 1: 99 | return tensors[0] 100 | return torch.cat(tensors, dim) 101 | 102 | 103 | def cat_boxlist(bboxes): 104 | """ 105 | Concatenates a list of BoxList (having the same image size) into a 106 | single BoxList 107 | 108 | Arguments: 109 | bboxes (list[BoxList]) 110 | """ 111 | assert isinstance(bboxes, (list, tuple)) 112 | assert all(isinstance(bbox, BoxList) for bbox in bboxes) 113 | 114 | size = bboxes[0].size 115 | assert all(bbox.size == size for bbox in bboxes) 116 | 117 | mode = bboxes[0].mode 118 | assert all(bbox.mode == mode for bbox in bboxes) 119 | 120 | fields = set(bboxes[0].fields()) 121 | assert all(set(bbox.fields()) == fields for bbox in bboxes) 122 | 123 | cat_boxes = BoxList(_cat([bbox.bbox for bbox in bboxes], dim=0), size, mode) 124 | 125 | for field in fields: 126 | data = _cat([bbox.get_field(field) for bbox in bboxes], dim=0) 127 | cat_boxes.add_field(field, data) 128 | 129 | return cat_boxes 130 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/backbone/fpn.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | import torch.nn.functional as F 4 | from torch import nn 5 | 6 | 7 | class FPN(nn.Module): 8 | """ 9 | Module that adds FPN on top of a list of feature maps. 10 | The feature maps are currently supposed to be in increasing depth 11 | order, and must be consecutive 12 | """ 13 | 14 | def __init__( 15 | self, in_channels_list, out_channels, conv_block, top_blocks=None 16 | ): 17 | """ 18 | Arguments: 19 | in_channels_list (list[int]): number of channels for each feature map that 20 | will be fed 21 | out_channels (int): number of channels of the FPN representation 22 | top_blocks (nn.Module or None): if provided, an extra operation will 23 | be performed on the output of the last (smallest resolution) 24 | FPN output, and the result will extend the result list 25 | """ 26 | super(FPN, self).__init__() 27 | self.inner_blocks = [] 28 | self.layer_blocks = [] 29 | for idx, in_channels in enumerate(in_channels_list, 1): 30 | inner_block = "fpn_inner{}".format(idx) 31 | layer_block = "fpn_layer{}".format(idx) 32 | 33 | if in_channels == 0: 34 | continue 35 | inner_block_module = conv_block(in_channels, out_channels, 1) 36 | layer_block_module = conv_block(out_channels, out_channels, 3, 1) 37 | self.add_module(inner_block, inner_block_module) 38 | self.add_module(layer_block, layer_block_module) 39 | self.inner_blocks.append(inner_block) 40 | self.layer_blocks.append(layer_block) 41 | self.top_blocks = top_blocks 42 | 43 | def forward(self, x): 44 | """ 45 | Arguments: 46 | x (list[Tensor]): feature maps for each feature level. 47 | Returns: 48 | results (tuple[Tensor]): feature maps after FPN layers. 49 | They are ordered from highest resolution first. 50 | """ 51 | last_inner = getattr(self, self.inner_blocks[-1])(x[-1]) 52 | results = [] 53 | results.append(getattr(self, self.layer_blocks[-1])(last_inner)) 54 | for feature, inner_block, layer_block in zip( 55 | x[:-1][::-1], self.inner_blocks[:-1][::-1], self.layer_blocks[:-1][::-1] 56 | ): 57 | if not inner_block: 58 | continue 59 | inner_top_down = F.interpolate(last_inner, scale_factor=2, mode="nearest") 60 | inner_lateral = getattr(self, inner_block)(feature) 61 | # TODO use size instead of scale to make it robust to different sizes 62 | # inner_top_down = F.upsample(last_inner, size=inner_lateral.shape[-2:], 63 | # mode='bilinear', align_corners=False) 64 | last_inner = inner_lateral + inner_top_down 65 | results.insert(0, getattr(self, layer_block)(last_inner)) 66 | 67 | if isinstance(self.top_blocks, LastLevelP6P7): 68 | last_results = self.top_blocks(x[-1], results[-1]) 69 | results.extend(last_results) 70 | elif isinstance(self.top_blocks, LastLevelMaxPool): 71 | last_results = self.top_blocks(results[-1]) 72 | results.extend(last_results) 73 | 74 | return tuple(results) 75 | 76 | 77 | class LastLevelMaxPool(nn.Module): 78 | def forward(self, x): 79 | return [F.max_pool2d(x, 1, 2, 0)] 80 | 81 | 82 | class LastLevelP6P7(nn.Module): 83 | """ 84 | This module is used in RetinaNet to generate extra layers, P6 and P7. 85 | """ 86 | def __init__(self, in_channels, out_channels): 87 | super(LastLevelP6P7, self).__init__() 88 | self.p6 = nn.Conv2d(in_channels, out_channels, 3, 2, 1) 89 | self.p7 = nn.Conv2d(out_channels, out_channels, 3, 2, 1) 90 | for module in [self.p6, self.p7]: 91 | nn.init.kaiming_uniform_(module.weight, a=1) 92 | nn.init.constant_(module.bias, 0) 93 | self.use_P5 = in_channels == out_channels 94 | 95 | def forward(self, c5, p5): 96 | x = p5 if self.use_P5 else c5 97 | p6 = self.p6(x) 98 | p7 = self.p7(F.relu(p6)) 99 | return [p6, p7] 100 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/engine/inference.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import logging 3 | import time 4 | import os 5 | 6 | import torch 7 | from tqdm import tqdm 8 | 9 | from maskrcnn_benchmark.config import cfg 10 | from maskrcnn_benchmark.data.datasets.evaluation import evaluate 11 | from ..utils.comm import is_main_process, get_world_size 12 | from ..utils.comm import all_gather 13 | from ..utils.comm import synchronize 14 | from ..utils.timer import Timer, get_time_str 15 | from .bbox_aug import im_detect_bbox_aug 16 | 17 | 18 | def compute_on_dataset(model, data_loader, device, timer=None): 19 | model.eval() 20 | results_dict = {} 21 | cpu_device = torch.device("cpu") 22 | for _, batch in enumerate(tqdm(data_loader)): 23 | images, targets, image_ids = batch 24 | with torch.no_grad(): 25 | if timer: 26 | timer.tic() 27 | if cfg.TEST.BBOX_AUG.ENABLED: 28 | output = im_detect_bbox_aug(model, images, device) 29 | else: 30 | output = model(images.to(device)) 31 | if timer: 32 | if not cfg.MODEL.DEVICE == 'cpu': 33 | torch.cuda.synchronize() 34 | timer.toc() 35 | output = [o.to(cpu_device) for o in output] 36 | results_dict.update( 37 | {img_id: result for img_id, result in zip(image_ids, output)} 38 | ) 39 | return results_dict 40 | 41 | 42 | def _accumulate_predictions_from_multiple_gpus(predictions_per_gpu): 43 | all_predictions = all_gather(predictions_per_gpu) 44 | if not is_main_process(): 45 | return 46 | # merge the list of dicts 47 | predictions = {} 48 | for p in all_predictions: 49 | predictions.update(p) 50 | # convert a dict where the key is the index in a list 51 | image_ids = list(sorted(predictions.keys())) 52 | if len(image_ids) != image_ids[-1] + 1: 53 | logger = logging.getLogger("maskrcnn_benchmark.inference") 54 | logger.warning( 55 | "Number of images that were gathered from multiple processes is not " 56 | "a contiguous set. Some images might be missing from the evaluation" 57 | ) 58 | 59 | # convert to a list 60 | predictions = [predictions[i] for i in image_ids] 61 | return predictions 62 | 63 | 64 | def inference( 65 | model, 66 | data_loader, 67 | dataset_name, 68 | iou_types=("bbox",), 69 | box_only=False, 70 | device="cuda", 71 | expected_results=(), 72 | expected_results_sigma_tol=4, 73 | output_folder=None, 74 | ): 75 | # convert to a torch.device for efficiency 76 | device = torch.device(device) 77 | num_devices = get_world_size() 78 | logger = logging.getLogger("maskrcnn_benchmark.inference") 79 | dataset = data_loader.dataset 80 | logger.info("Start evaluation on {} dataset({} images).".format(dataset_name, len(dataset))) 81 | total_timer = Timer() 82 | inference_timer = Timer() 83 | total_timer.tic() 84 | predictions = compute_on_dataset(model, data_loader, device, inference_timer) 85 | # wait for all processes to complete before measuring the time 86 | synchronize() 87 | total_time = total_timer.toc() 88 | total_time_str = get_time_str(total_time) 89 | logger.info( 90 | "Total run time: {} ({} s / img per device, on {} devices)".format( 91 | total_time_str, total_time * num_devices / len(dataset), num_devices 92 | ) 93 | ) 94 | total_infer_time = get_time_str(inference_timer.total_time) 95 | logger.info( 96 | "Model inference time: {} ({} s / img per device, on {} devices)".format( 97 | total_infer_time, 98 | inference_timer.total_time * num_devices / len(dataset), 99 | num_devices, 100 | ) 101 | ) 102 | 103 | predictions = _accumulate_predictions_from_multiple_gpus(predictions) 104 | if not is_main_process(): 105 | return 106 | 107 | if output_folder: 108 | torch.save(predictions, os.path.join(output_folder, "predictions.pth")) 109 | 110 | extra_args = dict( 111 | box_only=box_only, 112 | iou_types=iou_types, 113 | expected_results=expected_results, 114 | expected_results_sigma_tol=expected_results_sigma_tol, 115 | ) 116 | 117 | return evaluate(dataset=dataset, 118 | predictions=predictions, 119 | output_folder=output_folder, 120 | **extra_args) 121 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/datasets/voc.py: -------------------------------------------------------------------------------- 1 | import os 2 | 3 | import torch 4 | import torch.utils.data 5 | from PIL import Image 6 | import sys 7 | 8 | if sys.version_info[0] == 2: 9 | import xml.etree.cElementTree as ET 10 | else: 11 | import xml.etree.ElementTree as ET 12 | 13 | 14 | from maskrcnn_benchmark.structures.bounding_box import BoxList 15 | 16 | 17 | class PascalVOCDataset(torch.utils.data.Dataset): 18 | 19 | CLASSES = ( 20 | "__background__ ", 21 | "aeroplane", 22 | "bicycle", 23 | "bird", 24 | "boat", 25 | "bottle", 26 | "bus", 27 | "car", 28 | "cat", 29 | "chair", 30 | "cow", 31 | "diningtable", 32 | "dog", 33 | "horse", 34 | "motorbike", 35 | "person", 36 | "pottedplant", 37 | "sheep", 38 | "sofa", 39 | "train", 40 | "tvmonitor", 41 | ) 42 | 43 | def __init__(self, data_dir, split, use_difficult=False, transforms=None): 44 | self.root = data_dir 45 | self.image_set = split 46 | self.keep_difficult = use_difficult 47 | self.transforms = transforms 48 | 49 | self._annopath = os.path.join(self.root, "Annotations", "%s.xml") 50 | self._imgpath = os.path.join(self.root, "JPEGImages", "%s.jpg") 51 | self._imgsetpath = os.path.join(self.root, "ImageSets", "Main", "%s.txt") 52 | 53 | with open(self._imgsetpath % self.image_set) as f: 54 | self.ids = f.readlines() 55 | self.ids = [x.strip("\n") for x in self.ids] 56 | self.id_to_img_map = {k: v for k, v in enumerate(self.ids)} 57 | 58 | cls = PascalVOCDataset.CLASSES 59 | self.class_to_ind = dict(zip(cls, range(len(cls)))) 60 | self.categories = dict(zip(range(len(cls)), cls)) 61 | 62 | def __getitem__(self, index): 63 | img_id = self.ids[index] 64 | img = Image.open(self._imgpath % img_id).convert("RGB") 65 | 66 | target = self.get_groundtruth(index) 67 | target = target.clip_to_image(remove_empty=True) 68 | 69 | if self.transforms is not None: 70 | img, target = self.transforms(img, target) 71 | 72 | return img, target, index 73 | 74 | def __len__(self): 75 | return len(self.ids) 76 | 77 | def get_groundtruth(self, index): 78 | img_id = self.ids[index] 79 | anno = ET.parse(self._annopath % img_id).getroot() 80 | anno = self._preprocess_annotation(anno) 81 | 82 | height, width = anno["im_info"] 83 | target = BoxList(anno["boxes"], (width, height), mode="xyxy") 84 | target.add_field("labels", anno["labels"]) 85 | target.add_field("difficult", anno["difficult"]) 86 | return target 87 | 88 | def _preprocess_annotation(self, target): 89 | boxes = [] 90 | gt_classes = [] 91 | difficult_boxes = [] 92 | TO_REMOVE = 1 93 | 94 | for obj in target.iter("object"): 95 | difficult = int(obj.find("difficult").text) == 1 96 | if not self.keep_difficult and difficult: 97 | continue 98 | name = obj.find("name").text.lower().strip() 99 | bb = obj.find("bndbox") 100 | # Make pixel indexes 0-based 101 | # Refer to "https://github.com/rbgirshick/py-faster-rcnn/blob/master/lib/datasets/pascal_voc.py#L208-L211" 102 | box = [ 103 | bb.find("xmin").text, 104 | bb.find("ymin").text, 105 | bb.find("xmax").text, 106 | bb.find("ymax").text, 107 | ] 108 | bndbox = tuple( 109 | map(lambda x: x - TO_REMOVE, list(map(int, box))) 110 | ) 111 | 112 | boxes.append(bndbox) 113 | gt_classes.append(self.class_to_ind[name]) 114 | difficult_boxes.append(difficult) 115 | 116 | size = target.find("size") 117 | im_info = tuple(map(int, (size.find("height").text, size.find("width").text))) 118 | 119 | res = { 120 | "boxes": torch.tensor(boxes, dtype=torch.float32), 121 | "labels": torch.tensor(gt_classes), 122 | "difficult": torch.tensor(difficult_boxes), 123 | "im_info": im_info, 124 | } 125 | return res 126 | 127 | def get_img_info(self, index): 128 | img_id = self.ids[index] 129 | anno = ET.parse(self._annopath % img_id).getroot() 130 | size = anno.find("size") 131 | im_info = tuple(map(int, (size.find("height").text, size.find("width").text))) 132 | return {"height": im_info[0], "width": im_info[1]} 133 | 134 | def map_class_id_to_class_name(self, class_id): 135 | return PascalVOCDataset.CLASSES[class_id] 136 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/engine/trainer.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import datetime 3 | import logging 4 | import time 5 | 6 | import torch 7 | import torch.distributed as dist 8 | 9 | from maskrcnn_benchmark.utils.comm import get_world_size 10 | from maskrcnn_benchmark.utils.metric_logger import MetricLogger 11 | 12 | from apex import amp 13 | 14 | def reduce_loss_dict(loss_dict): 15 | """ 16 | Reduce the loss dictionary from all processes so that process with rank 17 | 0 has the averaged results. Returns a dict with the same fields as 18 | loss_dict, after reduction. 19 | """ 20 | world_size = get_world_size() 21 | if world_size < 2: 22 | return loss_dict 23 | with torch.no_grad(): 24 | loss_names = [] 25 | all_losses = [] 26 | for k in sorted(loss_dict.keys()): 27 | loss_names.append(k) 28 | all_losses.append(loss_dict[k]) 29 | all_losses = torch.stack(all_losses, dim=0) 30 | dist.reduce(all_losses, dst=0) 31 | if dist.get_rank() == 0: 32 | # only main process gets accumulated, so only divide by 33 | # world_size in this case 34 | all_losses /= world_size 35 | reduced_losses = {k: v for k, v in zip(loss_names, all_losses)} 36 | return reduced_losses 37 | 38 | 39 | def do_train( 40 | model, 41 | data_loader, 42 | optimizer, 43 | scheduler, 44 | checkpointer, 45 | device, 46 | checkpoint_period, 47 | arguments, 48 | ): 49 | logger = logging.getLogger("maskrcnn_benchmark.trainer") 50 | logger.info("Start training") 51 | meters = MetricLogger(delimiter=" ") 52 | max_iter = len(data_loader) 53 | start_iter = arguments["iteration"] 54 | model.train() 55 | start_training_time = time.time() 56 | end = time.time() 57 | for iteration, (images, targets, _) in enumerate(data_loader, start_iter): 58 | 59 | if any(len(target) < 1 for target in targets): 60 | logger.error(f"Iteration={iteration + 1} || Image Ids used for training {_} || targets Length={[len(target) for target in targets]}" ) 61 | continue 62 | data_time = time.time() - end 63 | iteration = iteration + 1 64 | arguments["iteration"] = iteration 65 | 66 | scheduler.step() 67 | 68 | images = images.to(device) 69 | targets = [target.to(device) for target in targets] 70 | 71 | loss_dict = model(images, targets) 72 | 73 | losses = sum(loss for loss in loss_dict.values()) 74 | 75 | # reduce losses over all GPUs for logging purposes 76 | loss_dict_reduced = reduce_loss_dict(loss_dict) 77 | losses_reduced = sum(loss for loss in loss_dict_reduced.values()) 78 | meters.update(loss=losses_reduced, **loss_dict_reduced) 79 | 80 | optimizer.zero_grad() 81 | # Note: If mixed precision is not used, this ends up doing nothing 82 | # Otherwise apply loss scaling for mixed-precision recipe 83 | with amp.scale_loss(losses, optimizer) as scaled_losses: 84 | scaled_losses.backward() 85 | optimizer.step() 86 | 87 | batch_time = time.time() - end 88 | end = time.time() 89 | meters.update(time=batch_time, data=data_time) 90 | 91 | eta_seconds = meters.time.global_avg * (max_iter - iteration) 92 | eta_string = str(datetime.timedelta(seconds=int(eta_seconds))) 93 | 94 | if iteration % 20 == 0 or iteration == max_iter: 95 | logger.info( 96 | meters.delimiter.join( 97 | [ 98 | "eta: {eta}", 99 | "iter: {iter}", 100 | "{meters}", 101 | "lr: {lr:.6f}", 102 | "max mem: {memory:.0f}", 103 | ] 104 | ).format( 105 | eta=eta_string, 106 | iter=iteration, 107 | meters=str(meters), 108 | lr=optimizer.param_groups[0]["lr"], 109 | memory=torch.cuda.max_memory_allocated() / 1024.0 / 1024.0, 110 | ) 111 | ) 112 | if iteration % checkpoint_period == 0: 113 | checkpointer.save("model_{:07d}".format(iteration), **arguments) 114 | if iteration == max_iter: 115 | checkpointer.save("model_final", **arguments) 116 | 117 | total_training_time = time.time() - start_training_time 118 | total_time_str = str(datetime.timedelta(seconds=total_training_time)) 119 | logger.info( 120 | "Total training time: {} ({:.4f} s / it)".format( 121 | total_time_str, total_training_time / (max_iter) 122 | ) 123 | ) 124 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/engine/bbox_aug.py: -------------------------------------------------------------------------------- 1 | import torch 2 | import torchvision.transforms as TT 3 | 4 | from maskrcnn_benchmark.config import cfg 5 | from maskrcnn_benchmark.data import transforms as T 6 | from maskrcnn_benchmark.structures.image_list import to_image_list 7 | from maskrcnn_benchmark.structures.bounding_box import BoxList 8 | from maskrcnn_benchmark.modeling.roi_heads.box_head.inference import make_roi_box_post_processor 9 | 10 | 11 | def im_detect_bbox_aug(model, images, device): 12 | # Collect detections computed under different transformations 13 | boxlists_ts = [] 14 | for _ in range(len(images)): 15 | boxlists_ts.append([]) 16 | 17 | def add_preds_t(boxlists_t): 18 | for i, boxlist_t in enumerate(boxlists_t): 19 | if len(boxlists_ts[i]) == 0: 20 | # The first one is identity transform, no need to resize the boxlist 21 | boxlists_ts[i].append(boxlist_t) 22 | else: 23 | # Resize the boxlist as the first one 24 | boxlists_ts[i].append(boxlist_t.resize(boxlists_ts[i][0].size)) 25 | 26 | # Compute detections for the original image (identity transform) 27 | boxlists_i = im_detect_bbox( 28 | model, images, cfg.INPUT.MIN_SIZE_TEST, cfg.INPUT.MAX_SIZE_TEST, device 29 | ) 30 | add_preds_t(boxlists_i) 31 | 32 | # Perform detection on the horizontally flipped image 33 | if cfg.TEST.BBOX_AUG.H_FLIP: 34 | boxlists_hf = im_detect_bbox_hflip( 35 | model, images, cfg.INPUT.MIN_SIZE_TEST, cfg.INPUT.MAX_SIZE_TEST, device 36 | ) 37 | add_preds_t(boxlists_hf) 38 | 39 | # Compute detections at different scales 40 | for scale in cfg.TEST.BBOX_AUG.SCALES: 41 | max_size = cfg.TEST.BBOX_AUG.MAX_SIZE 42 | boxlists_scl = im_detect_bbox_scale( 43 | model, images, scale, max_size, device 44 | ) 45 | add_preds_t(boxlists_scl) 46 | 47 | if cfg.TEST.BBOX_AUG.SCALE_H_FLIP: 48 | boxlists_scl_hf = im_detect_bbox_scale( 49 | model, images, scale, max_size, device, hflip=True 50 | ) 51 | add_preds_t(boxlists_scl_hf) 52 | 53 | # Merge boxlists detected by different bbox aug params 54 | boxlists = [] 55 | for i, boxlist_ts in enumerate(boxlists_ts): 56 | bbox = torch.cat([boxlist_t.bbox for boxlist_t in boxlist_ts]) 57 | scores = torch.cat([boxlist_t.get_field('scores') for boxlist_t in boxlist_ts]) 58 | boxlist = BoxList(bbox, boxlist_ts[0].size, boxlist_ts[0].mode) 59 | boxlist.add_field('scores', scores) 60 | boxlists.append(boxlist) 61 | 62 | # Apply NMS and limit the final detections 63 | results = [] 64 | post_processor = make_roi_box_post_processor(cfg) 65 | for boxlist in boxlists: 66 | results.append(post_processor.filter_results(boxlist, cfg.MODEL.ROI_BOX_HEAD.NUM_CLASSES)) 67 | 68 | return results 69 | 70 | 71 | def im_detect_bbox(model, images, target_scale, target_max_size, device): 72 | """ 73 | Performs bbox detection on the original image. 74 | """ 75 | transform = TT.Compose([ 76 | T.Resize(target_scale, target_max_size), 77 | TT.ToTensor(), 78 | T.Normalize( 79 | mean=cfg.INPUT.PIXEL_MEAN, std=cfg.INPUT.PIXEL_STD, to_bgr255=cfg.INPUT.TO_BGR255 80 | ) 81 | ]) 82 | images = [transform(image) for image in images] 83 | images = to_image_list(images, cfg.DATALOADER.SIZE_DIVISIBILITY) 84 | return model(images.to(device)) 85 | 86 | 87 | def im_detect_bbox_hflip(model, images, target_scale, target_max_size, device): 88 | """ 89 | Performs bbox detection on the horizontally flipped image. 90 | Function signature is the same as for im_detect_bbox. 91 | """ 92 | transform = TT.Compose([ 93 | T.Resize(target_scale, target_max_size), 94 | TT.RandomHorizontalFlip(1.0), 95 | TT.ToTensor(), 96 | T.Normalize( 97 | mean=cfg.INPUT.PIXEL_MEAN, std=cfg.INPUT.PIXEL_STD, to_bgr255=cfg.INPUT.TO_BGR255 98 | ) 99 | ]) 100 | images = [transform(image) for image in images] 101 | images = to_image_list(images, cfg.DATALOADER.SIZE_DIVISIBILITY) 102 | boxlists = model(images.to(device)) 103 | 104 | # Invert the detections computed on the flipped image 105 | boxlists_inv = [boxlist.transpose(0) for boxlist in boxlists] 106 | return boxlists_inv 107 | 108 | 109 | def im_detect_bbox_scale(model, images, target_scale, target_max_size, device, hflip=False): 110 | """ 111 | Computes bbox detections at the given scale. 112 | Returns predictions in the scaled image space. 113 | """ 114 | if hflip: 115 | boxlists_scl = im_detect_bbox_hflip(model, images, target_scale, target_max_size, device) 116 | else: 117 | boxlists_scl = im_detect_bbox(model, images, target_scale, target_max_size, device) 118 | return boxlists_scl 119 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/roi_heads/keypoint_head/inference.py: -------------------------------------------------------------------------------- 1 | import torch 2 | from torch import nn 3 | 4 | 5 | class KeypointPostProcessor(nn.Module): 6 | def __init__(self, keypointer=None): 7 | super(KeypointPostProcessor, self).__init__() 8 | self.keypointer = keypointer 9 | 10 | def forward(self, x, boxes): 11 | mask_prob = x 12 | 13 | scores = None 14 | if self.keypointer: 15 | mask_prob, scores = self.keypointer(x, boxes) 16 | 17 | assert len(boxes) == 1, "Only non-batched inference supported for now" 18 | boxes_per_image = [box.bbox.size(0) for box in boxes] 19 | mask_prob = mask_prob.split(boxes_per_image, dim=0) 20 | scores = scores.split(boxes_per_image, dim=0) 21 | 22 | results = [] 23 | for prob, box, score in zip(mask_prob, boxes, scores): 24 | bbox = BoxList(box.bbox, box.size, mode="xyxy") 25 | for field in box.fields(): 26 | bbox.add_field(field, box.get_field(field)) 27 | prob = PersonKeypoints(prob, box.size) 28 | prob.add_field("logits", score) 29 | bbox.add_field("keypoints", prob) 30 | results.append(bbox) 31 | 32 | return results 33 | 34 | 35 | # TODO remove and use only the Keypointer 36 | import numpy as np 37 | import cv2 38 | 39 | 40 | def heatmaps_to_keypoints(maps, rois): 41 | """Extract predicted keypoint locations from heatmaps. Output has shape 42 | (#rois, 4, #keypoints) with the 4 rows corresponding to (x, y, logit, prob) 43 | for each keypoint. 44 | """ 45 | # This function converts a discrete image coordinate in a HEATMAP_SIZE x 46 | # HEATMAP_SIZE image to a continuous keypoint coordinate. We maintain 47 | # consistency with keypoints_to_heatmap_labels by using the conversion from 48 | # Heckbert 1990: c = d + 0.5, where d is a discrete coordinate and c is a 49 | # continuous coordinate. 50 | offset_x = rois[:, 0] 51 | offset_y = rois[:, 1] 52 | 53 | widths = rois[:, 2] - rois[:, 0] 54 | heights = rois[:, 3] - rois[:, 1] 55 | widths = np.maximum(widths, 1) 56 | heights = np.maximum(heights, 1) 57 | widths_ceil = np.ceil(widths) 58 | heights_ceil = np.ceil(heights) 59 | 60 | # NCHW to NHWC for use with OpenCV 61 | maps = np.transpose(maps, [0, 2, 3, 1]) 62 | min_size = 0 # cfg.KRCNN.INFERENCE_MIN_SIZE 63 | num_keypoints = maps.shape[3] 64 | xy_preds = np.zeros((len(rois), 3, num_keypoints), dtype=np.float32) 65 | end_scores = np.zeros((len(rois), num_keypoints), dtype=np.float32) 66 | for i in range(len(rois)): 67 | if min_size > 0: 68 | roi_map_width = int(np.maximum(widths_ceil[i], min_size)) 69 | roi_map_height = int(np.maximum(heights_ceil[i], min_size)) 70 | else: 71 | roi_map_width = widths_ceil[i] 72 | roi_map_height = heights_ceil[i] 73 | width_correction = widths[i] / roi_map_width 74 | height_correction = heights[i] / roi_map_height 75 | roi_map = cv2.resize( 76 | maps[i], (roi_map_width, roi_map_height), interpolation=cv2.INTER_CUBIC 77 | ) 78 | # Bring back to CHW 79 | roi_map = np.transpose(roi_map, [2, 0, 1]) 80 | # roi_map_probs = scores_to_probs(roi_map.copy()) 81 | w = roi_map.shape[2] 82 | pos = roi_map.reshape(num_keypoints, -1).argmax(axis=1) 83 | x_int = pos % w 84 | y_int = (pos - x_int) // w 85 | # assert (roi_map_probs[k, y_int, x_int] == 86 | # roi_map_probs[k, :, :].max()) 87 | x = (x_int + 0.5) * width_correction 88 | y = (y_int + 0.5) * height_correction 89 | xy_preds[i, 0, :] = x + offset_x[i] 90 | xy_preds[i, 1, :] = y + offset_y[i] 91 | xy_preds[i, 2, :] = 1 92 | end_scores[i, :] = roi_map[np.arange(num_keypoints), y_int, x_int] 93 | 94 | return np.transpose(xy_preds, [0, 2, 1]), end_scores 95 | 96 | 97 | from maskrcnn_benchmark.structures.bounding_box import BoxList 98 | from maskrcnn_benchmark.structures.keypoint import PersonKeypoints 99 | 100 | 101 | class Keypointer(object): 102 | """ 103 | Projects a set of masks in an image on the locations 104 | specified by the bounding boxes 105 | """ 106 | 107 | def __init__(self, padding=0): 108 | self.padding = padding 109 | 110 | def __call__(self, masks, boxes): 111 | # TODO do this properly 112 | if isinstance(boxes, BoxList): 113 | boxes = [boxes] 114 | assert len(boxes) == 1 115 | 116 | result, scores = heatmaps_to_keypoints( 117 | masks.detach().cpu().numpy(), boxes[0].bbox.cpu().numpy() 118 | ) 119 | return torch.from_numpy(result).to(masks.device), torch.as_tensor(scores, device=masks.device) 120 | 121 | 122 | def make_roi_keypoint_post_processor(cfg): 123 | keypointer = Keypointer() 124 | keypoint_post_processor = KeypointPostProcessor(keypointer) 125 | return keypoint_post_processor 126 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/modeling/poolers.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import torch 3 | import torch.nn.functional as F 4 | from torch import nn 5 | 6 | from maskrcnn_benchmark.layers import ROIAlign 7 | 8 | from .utils import cat 9 | 10 | 11 | class LevelMapper(object): 12 | """Determine which FPN level each RoI in a set of RoIs should map to based 13 | on the heuristic in the FPN paper. 14 | """ 15 | 16 | def __init__(self, k_min, k_max, canonical_scale=224, canonical_level=4, eps=1e-6): 17 | """ 18 | Arguments: 19 | k_min (int) 20 | k_max (int) 21 | canonical_scale (int) 22 | canonical_level (int) 23 | eps (float) 24 | """ 25 | self.k_min = k_min 26 | self.k_max = k_max 27 | self.s0 = canonical_scale 28 | self.lvl0 = canonical_level 29 | self.eps = eps 30 | 31 | def __call__(self, boxlists): 32 | """ 33 | Arguments: 34 | boxlists (list[BoxList]) 35 | """ 36 | # Compute level ids 37 | s = torch.sqrt(cat([boxlist.area() for boxlist in boxlists])) 38 | 39 | # Eqn.(1) in FPN paper 40 | target_lvls = torch.floor(self.lvl0 + torch.log2(s / self.s0 + self.eps)) 41 | target_lvls = torch.clamp(target_lvls, min=self.k_min, max=self.k_max) 42 | return target_lvls.to(torch.int64) - self.k_min 43 | 44 | 45 | class Pooler(nn.Module): 46 | """ 47 | Pooler for Detection with or without FPN. 48 | It currently hard-code ROIAlign in the implementation, 49 | but that can be made more generic later on. 50 | Also, the requirement of passing the scales is not strictly necessary, as they 51 | can be inferred from the size of the feature map / size of original image, 52 | which is available thanks to the BoxList. 53 | """ 54 | 55 | def __init__(self, output_size, scales, sampling_ratio): 56 | """ 57 | Arguments: 58 | output_size (list[tuple[int]] or list[int]): output size for the pooled region 59 | scales (list[float]): scales for each Pooler 60 | sampling_ratio (int): sampling ratio for ROIAlign 61 | """ 62 | super(Pooler, self).__init__() 63 | poolers = [] 64 | for scale in scales: 65 | poolers.append( 66 | ROIAlign( 67 | output_size, spatial_scale=scale, sampling_ratio=sampling_ratio 68 | ) 69 | ) 70 | self.poolers = nn.ModuleList(poolers) 71 | self.output_size = output_size 72 | # get the levels in the feature map by leveraging the fact that the network always 73 | # downsamples by a factor of 2 at each level. 74 | lvl_min = -torch.log2(torch.tensor(scales[0], dtype=torch.float32)).item() 75 | lvl_max = -torch.log2(torch.tensor(scales[-1], dtype=torch.float32)).item() 76 | self.map_levels = LevelMapper(lvl_min, lvl_max) 77 | 78 | def convert_to_roi_format(self, boxes): 79 | concat_boxes = cat([b.bbox for b in boxes], dim=0) 80 | device, dtype = concat_boxes.device, concat_boxes.dtype 81 | ids = cat( 82 | [ 83 | torch.full((len(b), 1), i, dtype=dtype, device=device) 84 | for i, b in enumerate(boxes) 85 | ], 86 | dim=0, 87 | ) 88 | rois = torch.cat([ids, concat_boxes], dim=1) 89 | return rois 90 | 91 | def forward(self, x, boxes): 92 | """ 93 | Arguments: 94 | x (list[Tensor]): feature maps for each level 95 | boxes (list[BoxList]): boxes to be used to perform the pooling operation. 96 | Returns: 97 | result (Tensor) 98 | """ 99 | num_levels = len(self.poolers) 100 | rois = self.convert_to_roi_format(boxes) 101 | if num_levels == 1: 102 | return self.poolers[0](x[0], rois) 103 | 104 | levels = self.map_levels(boxes) 105 | 106 | num_rois = len(rois) 107 | num_channels = x[0].shape[1] 108 | output_size = self.output_size[0] 109 | 110 | dtype, device = x[0].dtype, x[0].device 111 | result = torch.zeros( 112 | (num_rois, num_channels, output_size, output_size), 113 | dtype=dtype, 114 | device=device, 115 | ) 116 | for level, (per_level_feature, pooler) in enumerate(zip(x, self.poolers)): 117 | idx_in_level = torch.nonzero(levels == level).squeeze(1) 118 | rois_per_level = rois[idx_in_level] 119 | result[idx_in_level] = pooler(per_level_feature, rois_per_level).to(dtype) 120 | 121 | return result 122 | 123 | 124 | def make_pooler(cfg, head_name): 125 | resolution = cfg.MODEL[head_name].POOLER_RESOLUTION 126 | scales = cfg.MODEL[head_name].POOLER_SCALES 127 | sampling_ratio = cfg.MODEL[head_name].POOLER_SAMPLING_RATIO 128 | pooler = Pooler( 129 | output_size=(resolution, resolution), 130 | scales=scales, 131 | sampling_ratio=sampling_ratio, 132 | ) 133 | return pooler 134 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/data/samplers/grouped_batch_sampler.py: -------------------------------------------------------------------------------- 1 | # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | import itertools 3 | 4 | import torch 5 | from torch.utils.data.sampler import BatchSampler 6 | from torch.utils.data.sampler import Sampler 7 | 8 | 9 | class GroupedBatchSampler(BatchSampler): 10 | """ 11 | Wraps another sampler to yield a mini-batch of indices. 12 | It enforces that elements from the same group should appear in groups of batch_size. 13 | It also tries to provide mini-batches which follows an ordering which is 14 | as close as possible to the ordering from the original sampler. 15 | 16 | Arguments: 17 | sampler (Sampler): Base sampler. 18 | batch_size (int): Size of mini-batch. 19 | drop_uneven (bool): If ``True``, the sampler will drop the batches whose 20 | size is less than ``batch_size`` 21 | 22 | """ 23 | 24 | def __init__(self, sampler, group_ids, batch_size, drop_uneven=False): 25 | if not isinstance(sampler, Sampler): 26 | raise ValueError( 27 | "sampler should be an instance of " 28 | "torch.utils.data.Sampler, but got sampler={}".format(sampler) 29 | ) 30 | self.sampler = sampler 31 | self.group_ids = torch.as_tensor(group_ids) 32 | assert self.group_ids.dim() == 1 33 | self.batch_size = batch_size 34 | self.drop_uneven = drop_uneven 35 | 36 | self.groups = torch.unique(self.group_ids).sort(0)[0] 37 | 38 | self._can_reuse_batches = False 39 | 40 | def _prepare_batches(self): 41 | dataset_size = len(self.group_ids) 42 | # get the sampled indices from the sampler 43 | sampled_ids = torch.as_tensor(list(self.sampler)) 44 | # potentially not all elements of the dataset were sampled 45 | # by the sampler (e.g., DistributedSampler). 46 | # construct a tensor which contains -1 if the element was 47 | # not sampled, and a non-negative number indicating the 48 | # order where the element was sampled. 49 | # for example. if sampled_ids = [3, 1] and dataset_size = 5, 50 | # the order is [-1, 1, -1, 0, -1] 51 | order = torch.full((dataset_size,), -1, dtype=torch.int64) 52 | order[sampled_ids] = torch.arange(len(sampled_ids)) 53 | 54 | # get a mask with the elements that were sampled 55 | mask = order >= 0 56 | 57 | # find the elements that belong to each individual cluster 58 | clusters = [(self.group_ids == i) & mask for i in self.groups] 59 | # get relative order of the elements inside each cluster 60 | # that follows the order from the sampler 61 | relative_order = [order[cluster] for cluster in clusters] 62 | # with the relative order, find the absolute order in the 63 | # sampled space 64 | permutation_ids = [s[s.sort()[1]] for s in relative_order] 65 | # permute each cluster so that they follow the order from 66 | # the sampler 67 | permuted_clusters = [sampled_ids[idx] for idx in permutation_ids] 68 | 69 | # splits each cluster in batch_size, and merge as a list of tensors 70 | splits = [c.split(self.batch_size) for c in permuted_clusters] 71 | merged = tuple(itertools.chain.from_iterable(splits)) 72 | 73 | # now each batch internally has the right order, but 74 | # they are grouped by clusters. Find the permutation between 75 | # different batches that brings them as close as possible to 76 | # the order that we have in the sampler. For that, we will consider the 77 | # ordering as coming from the first element of each batch, and sort 78 | # correspondingly 79 | first_element_of_batch = [t[0].item() for t in merged] 80 | # get and inverse mapping from sampled indices and the position where 81 | # they occur (as returned by the sampler) 82 | inv_sampled_ids_map = {v: k for k, v in enumerate(sampled_ids.tolist())} 83 | # from the first element in each batch, get a relative ordering 84 | first_index_of_batch = torch.as_tensor( 85 | [inv_sampled_ids_map[s] for s in first_element_of_batch] 86 | ) 87 | 88 | # permute the batches so that they approximately follow the order 89 | # from the sampler 90 | permutation_order = first_index_of_batch.sort(0)[1].tolist() 91 | # finally, permute the batches 92 | batches = [merged[i].tolist() for i in permutation_order] 93 | 94 | if self.drop_uneven: 95 | kept = [] 96 | for batch in batches: 97 | if len(batch) == self.batch_size: 98 | kept.append(batch) 99 | batches = kept 100 | return batches 101 | 102 | def __iter__(self): 103 | if self._can_reuse_batches: 104 | batches = self._batches 105 | self._can_reuse_batches = False 106 | else: 107 | batches = self._prepare_batches() 108 | self._batches = batches 109 | return iter(batches) 110 | 111 | def __len__(self): 112 | if not hasattr(self, "_batches"): 113 | self._batches = self._prepare_batches() 114 | self._can_reuse_batches = True 115 | return len(self._batches) 116 | -------------------------------------------------------------------------------- /maskrcnn_benchmark/csrc/deform_conv.h: -------------------------------------------------------------------------------- 1 | // Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved. 2 | #pragma once 3 | #include "cpu/vision.h" 4 | 5 | #ifdef WITH_CUDA 6 | #include "cuda/vision.h" 7 | #endif 8 | 9 | 10 | // Interface for Python 11 | int deform_conv_forward( 12 | at::Tensor input, 13 | at::Tensor weight, 14 | at::Tensor offset, 15 | at::Tensor output, 16 | at::Tensor columns, 17 | at::Tensor ones, 18 | int kW, 19 | int kH, 20 | int dW, 21 | int dH, 22 | int padW, 23 | int padH, 24 | int dilationW, 25 | int dilationH, 26 | int group, 27 | int deformable_group, 28 | int im2col_step) 29 | { 30 | if (input.type().is_cuda()) { 31 | #ifdef WITH_CUDA 32 | return deform_conv_forward_cuda( 33 | input, weight, offset, output, columns, ones, 34 | kW, kH, dW, dH, padW, padH, dilationW, dilationH, 35 | group, deformable_group, im2col_step 36 | ); 37 | #else 38 | AT_ERROR("Not compiled with GPU support"); 39 | #endif 40 | } 41 | AT_ERROR("Not implemented on the CPU"); 42 | } 43 | 44 | 45 | int deform_conv_backward_input( 46 | at::Tensor input, 47 | at::Tensor offset, 48 | at::Tensor gradOutput, 49 | at::Tensor gradInput, 50 | at::Tensor gradOffset, 51 | at::Tensor weight, 52 | at::Tensor columns, 53 | int kW, 54 | int kH, 55 | int dW, 56 | int dH, 57 | int padW, 58 | int padH, 59 | int dilationW, 60 | int dilationH, 61 | int group, 62 | int deformable_group, 63 | int im2col_step) 64 | { 65 | if (input.type().is_cuda()) { 66 | #ifdef WITH_CUDA 67 | return deform_conv_backward_input_cuda( 68 | input, offset, gradOutput, gradInput, gradOffset, weight, columns, 69 | kW, kH, dW, dH, padW, padH, dilationW, dilationH, 70 | group, deformable_group, im2col_step 71 | ); 72 | #else 73 | AT_ERROR("Not compiled with GPU support"); 74 | #endif 75 | } 76 | AT_ERROR("Not implemented on the CPU"); 77 | } 78 | 79 | 80 | int deform_conv_backward_parameters( 81 | at::Tensor input, 82 | at::Tensor offset, 83 | at::Tensor gradOutput, 84 | at::Tensor gradWeight, // at::Tensor gradBias, 85 | at::Tensor columns, 86 | at::Tensor ones, 87 | int kW, 88 | int kH, 89 | int dW, 90 | int dH, 91 | int padW, 92 | int padH, 93 | int dilationW, 94 | int dilationH, 95 | int group, 96 | int deformable_group, 97 | float scale, 98 | int im2col_step) 99 | { 100 | if (input.type().is_cuda()) { 101 | #ifdef WITH_CUDA 102 | return deform_conv_backward_parameters_cuda( 103 | input, offset, gradOutput, gradWeight, columns, ones, 104 | kW, kH, dW, dH, padW, padH, dilationW, dilationH, 105 | group, deformable_group, scale, im2col_step 106 | ); 107 | #else 108 | AT_ERROR("Not compiled with GPU support"); 109 | #endif 110 | } 111 | AT_ERROR("Not implemented on the CPU"); 112 | } 113 | 114 | 115 | void modulated_deform_conv_forward( 116 | at::Tensor input, 117 | at::Tensor weight, 118 | at::Tensor bias, 119 | at::Tensor ones, 120 | at::Tensor offset, 121 | at::Tensor mask, 122 | at::Tensor output, 123 | at::Tensor columns, 124 | int kernel_h, 125 | int kernel_w, 126 | const int stride_h, 127 | const int stride_w, 128 | const int pad_h, 129 | const int pad_w, 130 | const int dilation_h, 131 | const int dilation_w, 132 | const int group, 133 | const int deformable_group, 134 | const bool with_bias) 135 | { 136 | if (input.type().is_cuda()) { 137 | #ifdef WITH_CUDA 138 | return modulated_deform_conv_cuda_forward( 139 | input, weight, bias, ones, offset, mask, output, columns, 140 | kernel_h, kernel_w, stride_h, stride_w, 141 | pad_h, pad_w, dilation_h, dilation_w, 142 | group, deformable_group, with_bias 143 | ); 144 | #else 145 | AT_ERROR("Not compiled with GPU support"); 146 | #endif 147 | } 148 | AT_ERROR("Not implemented on the CPU"); 149 | } 150 | 151 | 152 | void modulated_deform_conv_backward( 153 | at::Tensor input, 154 | at::Tensor weight, 155 | at::Tensor bias, 156 | at::Tensor ones, 157 | at::Tensor offset, 158 | at::Tensor mask, 159 | at::Tensor columns, 160 | at::Tensor grad_input, 161 | at::Tensor grad_weight, 162 | at::Tensor grad_bias, 163 | at::Tensor grad_offset, 164 | at::Tensor grad_mask, 165 | at::Tensor grad_output, 166 | int kernel_h, 167 | int kernel_w, 168 | int stride_h, 169 | int stride_w, 170 | int pad_h, 171 | int pad_w, 172 | int dilation_h, 173 | int dilation_w, 174 | int group, 175 | int deformable_group, 176 | const bool with_bias) 177 | { 178 | if (input.type().is_cuda()) { 179 | #ifdef WITH_CUDA 180 | return modulated_deform_conv_cuda_backward( 181 | input, weight, bias, ones, offset, mask, columns, 182 | grad_input, grad_weight, grad_bias, grad_offset, grad_mask, grad_output, 183 | kernel_h, kernel_w, stride_h, stride_w, pad_h, pad_w, dilation_h, dilation_w, 184 | group, deformable_group, with_bias 185 | ); 186 | #else 187 | AT_ERROR("Not compiled with GPU support"); 188 | #endif 189 | } 190 | AT_ERROR("Not implemented on the CPU"); 191 | } --------------------------------------------------------------------------------