[Feature] New Codec for IPR, a wrapper for multi task losses (#1628)

Co-authored-by: Yining Li <liyining0712@gmail.com>

configs/body_2d_keypoint/integral_regression/coco/ipr_res50_8xb64-210e_coco-256x256.py

@@ -0,0 +1,139 @@
+_base_ = ['../../../_base_/default_runtime.py']
+
+# runtime
+train_cfg = dict(max_epochs=210, val_interval=10)
+
+# optimizer
+optim_wrapper = dict(optimizer=dict(
+    type='Adam',
+    lr=5e-4,
+))
+
+# learning policy
+param_scheduler = [
+    dict(
+        type='LinearLR', begin=0, end=500, start_factor=0.001,
+        by_epoch=False),  # warm-up
+    dict(
+        type='MultiStepLR',
+        begin=0,
+        end=210,
+        milestones=[170, 200],
+        gamma=0.1,
+        by_epoch=True)
+]
+
+# automatically scaling LR based on the actual training batch size
+auto_scale_lr = dict(base_batch_size=512)
+
+# hooks
+default_hooks = dict(checkpoint=dict(save_best='coco/AP', rule='greater'))
+
+# codec settings
+codec = dict(
+    type='IntegralRegressionLabel',
+    input_size=(256, 256),
+    heatmap_size=(64, 64),
+    sigma=2.0,
+    normalize=True)
+
+# model settings
+model = dict(
+    type='TopdownPoseEstimator',
+    data_preprocessor=dict(
+        type='PoseDataPreprocessor',
+        mean=[123.675, 116.28, 103.53],
+        std=[58.395, 57.12, 57.375],
+        bgr_to_rgb=True),
+    backbone=dict(
+        type='ResNet',
+        depth=50,
+    ),
+    head=dict(
+        type='DSNTHead',
+        in_channels=2048,
+        in_featuremap_size=(8, 8),
+        num_joints=17,
+        loss=dict(
+            type='MultipleLossWrapper',
+            losses=[
+                dict(type='SmoothL1Loss', use_target_weight=True),
+                dict(type='KeypointMSELoss', use_target_weight=True)
+            ]),
+        decoder=codec),
+    test_cfg=dict(
+        flip_test=True,
+        shift_coords=True,
+        shift_heatmap=True,
+    ),
+    init_cfg=dict(
+        type='Pretrained',
+        checkpoint='https://download.openmmlab.com/mmpose/'
+        'pretrain_models/td-hm_res50_8xb64-210e_coco-256x192.pth'))
+
+# base dataset settings
+dataset_type = 'CocoDataset'
+data_mode = 'topdown'
+data_root = 'data/coco/'
+
+file_client_args = dict(backend='disk')
+
+# pipelines
+train_pipeline = [
+    dict(type='LoadImage', file_client_args=file_client_args),
+    dict(type='GetBBoxCenterScale'),
+    dict(type='RandomFlip', direction='horizontal'),
+    dict(type='RandomHalfBody'),
+    dict(type='RandomBBoxTransform'),
+    dict(type='TopdownAffine', input_size=codec['input_size']),
+    dict(
+        type='GenerateTarget',
+        target_type='heatmap+keypoint_label',
+        encoder=codec),
+    dict(type='PackPoseInputs')
+]
+test_pipeline = [
+    dict(type='LoadImage', file_client_args=file_client_args),
+    dict(type='GetBBoxCenterScale'),
+    dict(type='TopdownAffine', input_size=codec['input_size']),
+    dict(type='PackPoseInputs')
+]
+
+# data loaders
+train_dataloader = dict(
+    batch_size=64,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=True),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_mode=data_mode,
+        ann_file='annotations/person_keypoints_train2017.json',
+        data_prefix=dict(img='train2017/'),
+        pipeline=train_pipeline,
+    ))
+val_dataloader = dict(
+    batch_size=32,
+    num_workers=2,
+    persistent_workers=True,
+    drop_last=False,
+    sampler=dict(type='DefaultSampler', shuffle=False, round_up=False),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_mode=data_mode,
+        ann_file='annotations/person_keypoints_val2017.json',
+        bbox_file=f'{data_root}person_detection_results/'
+        'COCO_val2017_detections_AP_H_56_person.json',
+        data_prefix=dict(img='val2017/'),
+        test_mode=True,
+        pipeline=test_pipeline,
+    ))
+test_dataloader = val_dataloader
+
+# evaluators
+val_evaluator = dict(
+    type='CocoMetric',
+    ann_file=f'{data_root}annotations/person_keypoints_val2017.json')
+test_evaluator = val_evaluator

configs/body_2d_keypoint/integral_regression/coco/ipr_res50_debias-8xb64-210e_coco-256x256.py

@@ -0,0 +1,138 @@
+_base_ = ['../../../_base_/default_runtime.py']
+
+# runtime
+train_cfg = dict(max_epochs=210, val_interval=10)
+
+# optimizer
+optim_wrapper = dict(optimizer=dict(
+    type='Adam',
+    lr=5e-4,
+))
+
+# learning policy
+param_scheduler = [
+    dict(
+        type='LinearLR', begin=0, end=500, start_factor=0.001,
+        by_epoch=False),  # warm-up
+    dict(
+        type='MultiStepLR',
+        begin=0,
+        end=210,
+        milestones=[170, 200],
+        gamma=0.1,
+        by_epoch=True)
+]
+
+# automatically scaling LR based on the actual training batch size
+auto_scale_lr = dict(base_batch_size=512)
+
+# hooks
+default_hooks = dict(checkpoint=dict(save_best='coco/AP', rule='greater'))
+
+# codec settings
+codec = dict(
+    type='IntegralRegressionLabel',
+    input_size=(256, 256),
+    heatmap_size=(64, 64),
+    sigma=2.0,
+    normalize=True)
+
+# model settings
+model = dict(
+    type='TopdownPoseEstimator',
+    data_preprocessor=dict(
+        type='PoseDataPreprocessor',
+        mean=[123.675, 116.28, 103.53],
+        std=[58.395, 57.12, 57.375],
+        bgr_to_rgb=True),
+    backbone=dict(type='ResNet', depth=50),
+    head=dict(
+        type='DSNTHead',
+        in_channels=2048,
+        in_featuremap_size=(8, 8),
+        num_joints=17,
+        debias=True,
+        beta=10.,
+        loss=dict(
+            type='MultipleLossWrapper',
+            losses=[
+                dict(type='SmoothL1Loss', use_target_weight=True),
+                dict(type='JSDiscretLoss', use_target_weight=True)
+            ]),
+        decoder=codec),
+    test_cfg=dict(
+        flip_test=True,
+        shift_coords=True,
+        shift_heatmap=True,
+    ),
+    init_cfg=dict(
+        type='Pretrained',
+        checkpoint='https://download.openmmlab.com/mmpose/'
+        'pretrain_models/td-hm_res50_8xb64-210e_coco-256x192.pth'))
+
+# base dataset settings
+dataset_type = 'CocoDataset'
+data_mode = 'topdown'
+data_root = 'data/coco/'
+
+file_client_args = dict(backend='disk')
+
+# pipelines
+train_pipeline = [
+    dict(type='LoadImage', file_client_args=file_client_args),
+    dict(type='GetBBoxCenterScale'),
+    dict(type='RandomFlip', direction='horizontal'),
+    dict(type='RandomHalfBody'),
+    dict(type='RandomBBoxTransform'),
+    dict(type='TopdownAffine', input_size=codec['input_size']),
+    dict(
+        type='GenerateTarget',
+        target_type='heatmap+keypoint_label',
+        encoder=codec),
+    dict(type='PackPoseInputs')
+]
+test_pipeline = [
+    dict(type='LoadImage', file_client_args=file_client_args),
+    dict(type='GetBBoxCenterScale'),
+    dict(type='TopdownAffine', input_size=codec['input_size']),
+    dict(type='PackPoseInputs')
+]
+
+# data loaders
+train_dataloader = dict(
+    batch_size=64,
+    num_workers=2,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=True),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_mode=data_mode,
+        ann_file='annotations/person_keypoints_train2017.json',
+        data_prefix=dict(img='train2017/'),
+        pipeline=train_pipeline,
+    ))
+val_dataloader = dict(
+    batch_size=32,
+    num_workers=2,
+    persistent_workers=True,
+    drop_last=False,
+    sampler=dict(type='DefaultSampler', shuffle=False, round_up=False),
+    dataset=dict(
+        type=dataset_type,
+        data_root=data_root,
+        data_mode=data_mode,
+        ann_file='annotations/person_keypoints_val2017.json',
+        bbox_file=f'{data_root}person_detection_results/'
+        'COCO_val2017_detections_AP_H_56_person.json',
+        data_prefix=dict(img='val2017/'),
+        test_mode=True,
+        pipeline=test_pipeline,
+    ))
+test_dataloader = val_dataloader
+
+# evaluators
+val_evaluator = dict(
+    type='CocoMetric',
+    ann_file=f'{data_root}annotations/person_keypoints_val2017.json')
+test_evaluator = val_evaluator