Skip to content

GitLab

Commit 37c8cebc authored by Sugon_ldc's avatar Sugon_ldc
Browse files

add new model

parents
Pipeline #318 failed with stages
in 0 seconds
Hide whitespace changes
Inline Side-by-side
Showing with 2214 additions and 0 deletions
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/hrnet_w48_animalpose_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/animalpose.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=20,
dataset_joints=20,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
],
inference_channel=[
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19
])
# model settings
model = dict(
type='TopDown',
pretrained='https://download.openmmlab.com/mmpose/'
'pretrain_models/hrnet_w48-8ef0771d.pth',
backbone=dict(
type='HRNet',
in_channels=3,
extra=dict(
stage1=dict(
num_modules=1,
num_branches=1,
block='BOTTLENECK',
num_blocks=(4, ),
num_channels=(64, )),
stage2=dict(
num_modules=1,
num_branches=2,
block='BASIC',
num_blocks=(4, 4),
num_channels=(48, 96)),
stage3=dict(
num_modules=4,
num_branches=3,
block='BASIC',
num_blocks=(4, 4, 4),
num_channels=(48, 96, 192)),
stage4=dict(
num_modules=3,
num_branches=4,
block='BASIC',
num_blocks=(4, 4, 4, 4),
num_channels=(48, 96, 192, 384))),
),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=48,
out_channels=channel_cfg['num_output_channels'],
num_deconv_layers=0,
extra=dict(final_conv_kernel=1, ),
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.16, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/animalpose'
data = dict(
samples_per_gpu=64,
workers_per_gpu=2,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_train.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_val.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_val.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=test_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/res101_animalpose_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/animalpose.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=20,
dataset_joints=20,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
],
inference_channel=[
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19
])
# model settings
model = dict(
type='TopDown',
pretrained='torchvision://resnet101',
backbone=dict(type='ResNet', depth=101),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=2048,
out_channels=channel_cfg['num_output_channels'],
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.16, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/animalpose'
data = dict(
samples_per_gpu=64,
workers_per_gpu=2,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_train.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_val.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_val.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=test_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/res152_animalpose_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/animalpose.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=20,
dataset_joints=20,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
],
inference_channel=[
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19
])
# model settings
model = dict(
type='TopDown',
pretrained='torchvision://resnet152',
backbone=dict(type='ResNet', depth=152),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=2048,
out_channels=channel_cfg['num_output_channels'],
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.16, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/animalpose'
data = dict(
samples_per_gpu=32,
workers_per_gpu=2,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_train.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_val.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_val.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=test_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/res50_animalpose_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/animalpose.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=20,
dataset_joints=20,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
],
inference_channel=[
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19
])
# model settings
model = dict(
type='TopDown',
pretrained='torchvision://resnet50',
backbone=dict(type='ResNet', depth=50),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=2048,
out_channels=channel_cfg['num_output_channels'],
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.16, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/animalpose'
data = dict(
samples_per_gpu=64,
workers_per_gpu=2,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_train.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_val.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalPoseDataset',
ann_file=f'{data_root}/annotations/animalpose_val.json',
img_prefix=f'{data_root}/',
data_cfg=data_cfg,
pipeline=test_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/resnet_animalpose.md 0 → 100755
View file @ 37c8cebc
<!-- [ALGORITHM] -->
<details>
<summary align="right"><a href="http://openaccess.thecvf.com/content_ECCV_2018/html/Bin_Xiao_Simple_Baselines_for_ECCV_2018_paper.html">SimpleBaseline2D (ECCV'2018)</a></summary>
```bibtex
@inproceedings{xiao2018simple,
title={Simple baselines for human pose estimation and tracking},
author={Xiao, Bin and Wu, Haiping and Wei, Yichen},
booktitle={Proceedings of the European conference on computer vision (ECCV)},
pages={466--481},
year={2018}
}
```
</details>
<!-- [DATASET] -->
<details>
<summary align="right"><a href="http://openaccess.thecvf.com/content_ICCV_2019/html/Cao_Cross-Domain_Adaptation_for_Animal_Pose_Estimation_ICCV_2019_paper.html">Animal-Pose (ICCV'2019)</a></summary>
```bibtex
@InProceedings{Cao_2019_ICCV,
author = {Cao, Jinkun and Tang, Hongyang and Fang, Hao-Shu and Shen, Xiaoyong and Lu, Cewu and Tai, Yu-Wing},
title = {Cross-Domain Adaptation for Animal Pose Estimation},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {October},
year = {2019}
}
```
</details>
Results on AnimalPose validation set (1117 instances)
| Arch | Input Size | AP | AP<sup>50</sup> | AP<sup>75</sup> | AR | AR<sup>50</sup> | ckpt | log |
| :-------------------------------------------- | :--------: | :---: | :-------------: | :-------------: | :---: | :-------------: | :-------------------------------------------: | :-------------------------------------------: |
| [pose_resnet_50](/configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/res50_animalpose_256x256.py) | 256x256 | 0.688 | 0.945 | 0.772 | 0.733 | 0.952 | [ckpt](https://download.openmmlab.com/mmpose/animal/resnet/res50_animalpose_256x256-e1f30bff_20210426.pth) | [log](https://download.openmmlab.com/mmpose/animal/resnet/res50_animalpose_256x256_20210426.log.json) |
| [pose_resnet_101](/configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/res101_animalpose_256x256.py) | 256x256 | 0.696 | 0.948 | 0.785 | 0.737 | 0.954 | [ckpt](https://download.openmmlab.com/mmpose/animal/resnet/res101_animalpose_256x256-85563f4a_20210426.pth) | [log](https://download.openmmlab.com/mmpose/animal/resnet/res101_animalpose_256x256_20210426.log.json) |
| [pose_resnet_152](/configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/res152_animalpose_256x256.py) | 256x256 | 0.709 | 0.948 | 0.797 | 0.749 | 0.951 | [ckpt](https://download.openmmlab.com/mmpose/animal/resnet/res152_animalpose_256x256-a0a7506c_20210426.pth) | [log](https://download.openmmlab.com/mmpose/animal/resnet/res152_animalpose_256x256_20210426.log.json) |
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/resnet_animalpose.yml 0 → 100755
View file @ 37c8cebc
Collections:
- Name: SimpleBaseline2D
Paper:
Title: Simple baselines for human pose estimation and tracking
URL: http://openaccess.thecvf.com/content_ECCV_2018/html/Bin_Xiao_Simple_Baselines_for_ECCV_2018_paper.html
README: https://github.com/open-mmlab/mmpose/blob/master/docs/en/papers/algorithms/simplebaseline2d.md
Models:
- Config: configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/res50_animalpose_256x256.py
In Collection: SimpleBaseline2D
Metadata:
Architecture: &id001
- SimpleBaseline2D
Training Data: Animal-Pose
Name: topdown_heatmap_res50_animalpose_256x256
Results:
- Dataset: Animal-Pose
Metrics:
AP: 0.688
AP@0.5: 0.945
AP@0.75: 0.772
AR: 0.733
AR@0.5: 0.952
Task: Animal 2D Keypoint
Weights: https://download.openmmlab.com/mmpose/animal/resnet/res50_animalpose_256x256-e1f30bff_20210426.pth
- Config: configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/res101_animalpose_256x256.py
In Collection: SimpleBaseline2D
Metadata:
Architecture: *id001
Training Data: Animal-Pose
Name: topdown_heatmap_res101_animalpose_256x256
Results:
- Dataset: Animal-Pose
Metrics:
AP: 0.696
AP@0.5: 0.948
AP@0.75: 0.785
AR: 0.737
AR@0.5: 0.954
Task: Animal 2D Keypoint
Weights: https://download.openmmlab.com/mmpose/animal/resnet/res101_animalpose_256x256-85563f4a_20210426.pth
- Config: configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/animalpose/res152_animalpose_256x256.py
In Collection: SimpleBaseline2D
Metadata:
Architecture: *id001
Training Data: Animal-Pose
Name: topdown_heatmap_res152_animalpose_256x256
Results:
- Dataset: Animal-Pose
Metrics:
AP: 0.709
AP@0.5: 0.948
AP@0.75: 0.797
AR: 0.749
AR@0.5: 0.951
Task: Animal 2D Keypoint
Weights: https://download.openmmlab.com/mmpose/animal/resnet/res152_animalpose_256x256-a0a7506c_20210426.pth
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/hrnet_ap10k.md 0 → 100755
View file @ 37c8cebc
<!-- [ALGORITHM] -->
<details>
<summary align="right"><a href="http://openaccess.thecvf.com/content_CVPR_2019/html/Sun_Deep_High-Resolution_Representation_Learning_for_Human_Pose_Estimation_CVPR_2019_paper.html">HRNet (CVPR'2019)</a></summary>
```bibtex
@inproceedings{sun2019deep,
title={Deep high-resolution representation learning for human pose estimation},
author={Sun, Ke and Xiao, Bin and Liu, Dong and Wang, Jingdong},
booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
pages={5693--5703},
year={2019}
}
```
</details>
<!-- [DATASET] -->
<details>
<summary align="right"><a href="https://arxiv.org/abs/2108.12617">AP-10K (NeurIPS'2021)</a></summary>
```bibtex
@misc{yu2021ap10k,
title={AP-10K: A Benchmark for Animal Pose Estimation in the Wild},
author={Hang Yu and Yufei Xu and Jing Zhang and Wei Zhao and Ziyu Guan and Dacheng Tao},
year={2021},
eprint={2108.12617},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
</details>
Results on AP-10K validation set
| Arch | Input Size | AP | AP<sup>50</sup> | AP<sup>75</sup> | AP<sup>M</sup> | AP<sup>L</sup> | ckpt | log |
| :----------------------------------------- | :--------: | :---: | :-------------: | :-------------: | :------------: | :------------: | :-----------------------------------------: | :----------------------------------------: |
| [pose_hrnet_w32](/configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/hrnet_w32_ap10k_256x256.py) | 256x256 | 0.722 | 0.939 | 0.787 | 0.555 | 0.730 | [ckpt](https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w32_ap10k_256x256-18aac840_20211029.pth) | [log](https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w32_ap10k_256x256-18aac840_20211029.log.json) |
| [pose_hrnet_w48](/configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/hrnet_w48_ap10k_256x256.py) | 256x256 | 0.731 | 0.937 | 0.804 | 0.574 | 0.738 | [ckpt](https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w48_ap10k_256x256-d95ab412_20211029.pth) | [log](https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w48_ap10k_256x256-d95ab412_20211029.log.json) |
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/hrnet_ap10k.yml 0 → 100755
View file @ 37c8cebc
Collections:
- Name: HRNet
Paper:
Title: Deep high-resolution representation learning for human pose estimation
URL: http://openaccess.thecvf.com/content_CVPR_2019/html/Sun_Deep_High-Resolution_Representation_Learning_for_Human_Pose_Estimation_CVPR_2019_paper.html
README: https://github.com/open-mmlab/mmpose/blob/master/docs/en/papers/backbones/hrnet.md
Models:
- Config: configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/hrnet_w32_ap10k_256x256.py
In Collection: HRNet
Metadata:
Architecture: &id001
- HRNet
Training Data: AP-10K
Name: topdown_heatmap_hrnet_w32_ap10k_256x256
Results:
- Dataset: AP-10K
Metrics:
AP: 0.722
AP@0.5: 0.939
AP@0.75: 0.787
APL: 0.73
APM: 0.555
Task: Animal 2D Keypoint
Weights: https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w32_ap10k_256x256-18aac840_20211029.pth
- Config: configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/hrnet_w48_ap10k_256x256.py
In Collection: HRNet
Metadata:
Architecture: *id001
Training Data: AP-10K
Name: topdown_heatmap_hrnet_w48_ap10k_256x256
Results:
- Dataset: AP-10K
Metrics:
AP: 0.731
AP@0.5: 0.937
AP@0.75: 0.804
APL: 0.738
APM: 0.574
Task: Animal 2D Keypoint
Weights: https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w48_ap10k_256x256-d95ab412_20211029.pth
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/hrnet_w32_ap10k_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/ap10k.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=17,
dataset_joints=17,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16],
],
inference_channel=[
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16
])
# model settings
model = dict(
type='TopDown',
pretrained='https://download.openmmlab.com/mmpose/'
'pretrain_models/hrnet_w32-36af842e.pth',
backbone=dict(
type='HRNet',
in_channels=3,
extra=dict(
stage1=dict(
num_modules=1,
num_branches=1,
block='BOTTLENECK',
num_blocks=(4, ),
num_channels=(64, )),
stage2=dict(
num_modules=1,
num_branches=2,
block='BASIC',
num_blocks=(4, 4),
num_channels=(32, 64)),
stage3=dict(
num_modules=4,
num_branches=3,
block='BASIC',
num_blocks=(4, 4, 4),
num_channels=(32, 64, 128)),
stage4=dict(
num_modules=3,
num_branches=4,
block='BASIC',
num_blocks=(4, 4, 4, 4),
num_channels=(32, 64, 128, 256))),
),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=32,
out_channels=channel_cfg['num_output_channels'],
num_deconv_layers=0,
extra=dict(final_conv_kernel=1, ),
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.16, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/ap10k'
data = dict(
samples_per_gpu=64,
workers_per_gpu=4,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-train-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-val-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-test-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/hrnet_w48_ap10k_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/ap10k.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=17,
dataset_joints=17,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16],
],
inference_channel=[
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16
])
# model settings
model = dict(
type='TopDown',
pretrained='https://download.openmmlab.com/mmpose/'
'pretrain_models/hrnet_w48-8ef0771d.pth',
backbone=dict(
type='HRNet',
in_channels=3,
extra=dict(
stage1=dict(
num_modules=1,
num_branches=1,
block='BOTTLENECK',
num_blocks=(4, ),
num_channels=(64, )),
stage2=dict(
num_modules=1,
num_branches=2,
block='BASIC',
num_blocks=(4, 4),
num_channels=(48, 96)),
stage3=dict(
num_modules=4,
num_branches=3,
block='BASIC',
num_blocks=(4, 4, 4),
num_channels=(48, 96, 192)),
stage4=dict(
num_modules=3,
num_branches=4,
block='BASIC',
num_blocks=(4, 4, 4, 4),
num_channels=(48, 96, 192, 384))),
),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=48,
out_channels=channel_cfg['num_output_channels'],
num_deconv_layers=0,
extra=dict(final_conv_kernel=1, ),
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.16, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/ap10k'
data = dict(
samples_per_gpu=64,
workers_per_gpu=4,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-train-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-val-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-test-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/res101_ap10k_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/ap10k.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=17,
dataset_joints=17,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16],
],
inference_channel=[
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16
])
# model settings
model = dict(
type='TopDown',
pretrained='torchvision://resnet101',
backbone=dict(type='ResNet', depth=101),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=2048,
out_channels=channel_cfg['num_output_channels'],
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.16, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/ap10k'
data = dict(
samples_per_gpu=64,
workers_per_gpu=4,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-train-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-val-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-test-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/res50_ap10k_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/ap10k.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=17,
dataset_joints=17,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16],
],
inference_channel=[
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16
])
# model settings
model = dict(
type='TopDown',
pretrained='torchvision://resnet50',
backbone=dict(type='ResNet', depth=50),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=2048,
out_channels=channel_cfg['num_output_channels'],
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.16, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.25),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/ap10k'
data = dict(
samples_per_gpu=64,
workers_per_gpu=4,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-train-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-val-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalAP10KDataset',
ann_file=f'{data_root}/annotations/ap10k-test-split1.json',
img_prefix=f'{data_root}/data/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/resnet_ap10k.md 0 → 100755
View file @ 37c8cebc
<!-- [ALGORITHM] -->
<details>
<summary align="right"><a href="http://openaccess.thecvf.com/content_ECCV_2018/html/Bin_Xiao_Simple_Baselines_for_ECCV_2018_paper.html">SimpleBaseline2D (ECCV'2018)</a></summary>
```bibtex
@inproceedings{xiao2018simple,
title={Simple baselines for human pose estimation and tracking},
author={Xiao, Bin and Wu, Haiping and Wei, Yichen},
booktitle={Proceedings of the European conference on computer vision (ECCV)},
pages={466--481},
year={2018}
}
```
</details>
<!-- [DATASET] -->
<details>
<summary align="right"><a href="https://arxiv.org/abs/2108.12617">AP-10K (NeurIPS'2021)</a></summary>
```bibtex
@misc{yu2021ap10k,
title={AP-10K: A Benchmark for Animal Pose Estimation in the Wild},
author={Hang Yu and Yufei Xu and Jing Zhang and Wei Zhao and Ziyu Guan and Dacheng Tao},
year={2021},
eprint={2108.12617},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
```
</details>
Results on AP-10K validation set
| Arch | Input Size | AP | AP<sup>50</sup> | AP<sup>75</sup> | AP<sup>M</sup> | AP<sup>L</sup> | ckpt | log |
| :----------------------------------------- | :--------: | :---: | :-------------: | :-------------: | :------------: | :------------: | :-----------------------------------------: | :----------------------------------------: |
| [pose_resnet_50](/configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/res50_ap10k_256x256.py) | 256x256 | 0.681 | 0.923 | 0.740 | 0.510 | 0.688 | [ckpt](https://download.openmmlab.com/mmpose/animal/resnet/res50_ap10k_256x256-35760eb8_20211029.pth) | [log](https://download.openmmlab.com/mmpose/animal/resnet/res50_ap10k_256x256-35760eb8_20211029.log.json) |
| [pose_resnet_101](/configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/res101_ap10k_256x256.py) | 256x256 | 0.681 | 0.922 | 0.742 | 0.534 | 0.688 | [ckpt](https://download.openmmlab.com/mmpose/animal/resnet/res101_ap10k_256x256-9edfafb9_20211029.pth) | [log](https://download.openmmlab.com/mmpose/animal/resnet/res101_ap10k_256x256-9edfafb9_20211029.log.json) |
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/resnet_ap10k.yml 0 → 100755
View file @ 37c8cebc
Collections:
- Name: SimpleBaseline2D
Paper:
Title: Simple baselines for human pose estimation and tracking
URL: http://openaccess.thecvf.com/content_ECCV_2018/html/Bin_Xiao_Simple_Baselines_for_ECCV_2018_paper.html
README: https://github.com/open-mmlab/mmpose/blob/master/docs/en/papers/algorithms/simplebaseline2d.md
Models:
- Config: configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/res50_ap10k_256x256.py
In Collection: SimpleBaseline2D
Metadata:
Architecture: &id001
- SimpleBaseline2D
Training Data: AP-10K
Name: topdown_heatmap_res50_ap10k_256x256
Results:
- Dataset: AP-10K
Metrics:
AP: 0.681
AP@0.5: 0.923
AP@0.75: 0.74
APL: 0.688
APM: 0.51
Task: Animal 2D Keypoint
Weights: https://download.openmmlab.com/mmpose/animal/resnet/res50_ap10k_256x256-35760eb8_20211029.pth
- Config: configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/ap10k/res101_ap10k_256x256.py
In Collection: SimpleBaseline2D
Metadata:
Architecture: *id001
Training Data: AP-10K
Name: topdown_heatmap_res101_ap10k_256x256
Results:
- Dataset: AP-10K
Metrics:
AP: 0.681
AP@0.5: 0.922
AP@0.75: 0.742
APL: 0.688
APM: 0.534
Task: Animal 2D Keypoint
Weights: https://download.openmmlab.com/mmpose/animal/resnet/res101_ap10k_256x256-9edfafb9_20211029.pth
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/hrnet_atrw.md 0 → 100755
View file @ 37c8cebc
<!-- [ALGORITHM] -->
<details>
<summary align="right"><a href="http://openaccess.thecvf.com/content_CVPR_2019/html/Sun_Deep_High-Resolution_Representation_Learning_for_Human_Pose_Estimation_CVPR_2019_paper.html">HRNet (CVPR'2019)</a></summary>
```bibtex
@inproceedings{sun2019deep,
title={Deep high-resolution representation learning for human pose estimation},
author={Sun, Ke and Xiao, Bin and Liu, Dong and Wang, Jingdong},
booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
pages={5693--5703},
year={2019}
}
```
</details>
<!-- [DATASET] -->
<details>
<summary align="right"><a href="https://arxiv.org/abs/1906.05586">ATRW (ACM MM'2020)</a></summary>
```bibtex
@inproceedings{li2020atrw,
title={ATRW: A Benchmark for Amur Tiger Re-identification in the Wild},
author={Li, Shuyuan and Li, Jianguo and Tang, Hanlin and Qian, Rui and Lin, Weiyao},
booktitle={Proceedings of the 28th ACM International Conference on Multimedia},
pages={2590--2598},
year={2020}
}
```
</details>
Results on ATRW validation set
| Arch | Input Size | AP | AP<sup>50</sup> | AP<sup>75</sup> | AR | AR<sup>50</sup> | ckpt | log |
| :-------------------------------------------- | :--------: | :---: | :-------------: | :-------------: | :---: | :-------------: | :-------------------------------------------: | :-------------------------------------------: |
| [pose_hrnet_w32](/configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/hrnet_w32_atrw_256x256.py) | 256x256 | 0.912 | 0.973 | 0.959 | 0.938 | 0.985 | [ckpt](https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w32_atrw_256x256-f027f09a_20210414.pth) | [log](https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w32_atrw_256x256_20210414.log.json) |
| [pose_hrnet_w48](/configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/hrnet_w48_atrw_256x256.py) | 256x256 | 0.911 | 0.972 | 0.946 | 0.937 | 0.985 | [ckpt](https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w48_atrw_256x256-ac088892_20210414.pth) | [log](https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w48_atrw_256x256_20210414.log.json) |
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/hrnet_atrw.yml 0 → 100755
View file @ 37c8cebc
Collections:
- Name: HRNet
Paper:
Title: Deep high-resolution representation learning for human pose estimation
URL: http://openaccess.thecvf.com/content_CVPR_2019/html/Sun_Deep_High-Resolution_Representation_Learning_for_Human_Pose_Estimation_CVPR_2019_paper.html
README: https://github.com/open-mmlab/mmpose/blob/master/docs/en/papers/backbones/hrnet.md
Models:
- Config: configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/hrnet_w32_atrw_256x256.py
In Collection: HRNet
Metadata:
Architecture: &id001
- HRNet
Training Data: ATRW
Name: topdown_heatmap_hrnet_w32_atrw_256x256
Results:
- Dataset: ATRW
Metrics:
AP: 0.912
AP@0.5: 0.973
AP@0.75: 0.959
AR: 0.938
AR@0.5: 0.985
Task: Animal 2D Keypoint
Weights: https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w32_atrw_256x256-f027f09a_20210414.pth
- Config: configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/hrnet_w48_atrw_256x256.py
In Collection: HRNet
Metadata:
Architecture: *id001
Training Data: ATRW
Name: topdown_heatmap_hrnet_w48_atrw_256x256
Results:
- Dataset: ATRW
Metrics:
AP: 0.911
AP@0.5: 0.972
AP@0.75: 0.946
AR: 0.937
AR@0.5: 0.985
Task: Animal 2D Keypoint
Weights: https://download.openmmlab.com/mmpose/animal/hrnet/hrnet_w48_atrw_256x256-ac088892_20210414.pth
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/hrnet_w32_atrw_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/atrw.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=15,
dataset_joints=15,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14],
],
inference_channel=[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14])
# model settings
model = dict(
type='TopDown',
pretrained='https://download.openmmlab.com/mmpose/'
'pretrain_models/hrnet_w32-36af842e.pth',
backbone=dict(
type='HRNet',
in_channels=3,
extra=dict(
stage1=dict(
num_modules=1,
num_branches=1,
block='BOTTLENECK',
num_blocks=(4, ),
num_channels=(64, )),
stage2=dict(
num_modules=1,
num_branches=2,
block='BASIC',
num_blocks=(4, 4),
num_channels=(32, 64)),
stage3=dict(
num_modules=4,
num_branches=3,
block='BASIC',
num_blocks=(4, 4, 4),
num_channels=(32, 64, 128)),
stage4=dict(
num_modules=3,
num_branches=4,
block='BASIC',
num_blocks=(4, 4, 4, 4),
num_channels=(32, 64, 128, 256))),
),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=32,
out_channels=channel_cfg['num_output_channels'],
num_deconv_layers=0,
extra=dict(final_conv_kernel=1, ),
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.0),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.2, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.0),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/atrw'
data = dict(
samples_per_gpu=64,
workers_per_gpu=2,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_train.json',
img_prefix=f'{data_root}/images/train/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_val.json',
img_prefix=f'{data_root}/images/val/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_val.json',
img_prefix=f'{data_root}/images/val/',
data_cfg=data_cfg,
pipeline=test_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/hrnet_w48_atrw_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/atrw.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=15,
dataset_joints=15,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14],
],
inference_channel=[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14])
# model settings
model = dict(
type='TopDown',
pretrained='https://download.openmmlab.com/mmpose/'
'pretrain_models/hrnet_w48-8ef0771d.pth',
backbone=dict(
type='HRNet',
in_channels=3,
extra=dict(
stage1=dict(
num_modules=1,
num_branches=1,
block='BOTTLENECK',
num_blocks=(4, ),
num_channels=(64, )),
stage2=dict(
num_modules=1,
num_branches=2,
block='BASIC',
num_blocks=(4, 4),
num_channels=(48, 96)),
stage3=dict(
num_modules=4,
num_branches=3,
block='BASIC',
num_blocks=(4, 4, 4),
num_channels=(48, 96, 192)),
stage4=dict(
num_modules=3,
num_branches=4,
block='BASIC',
num_blocks=(4, 4, 4, 4),
num_channels=(48, 96, 192, 384))),
),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=48,
out_channels=channel_cfg['num_output_channels'],
num_deconv_layers=0,
extra=dict(final_conv_kernel=1, ),
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.0),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.2, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.0),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/atrw'
data = dict(
samples_per_gpu=64,
workers_per_gpu=2,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_train.json',
img_prefix=f'{data_root}/images/train/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_val.json',
img_prefix=f'{data_root}/images/val/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_val.json',
img_prefix=f'{data_root}/images/val/',
data_cfg=data_cfg,
pipeline=test_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/res101_atrw_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/atrw.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=15,
dataset_joints=15,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14],
],
inference_channel=[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14])
# model settings
model = dict(
type='TopDown',
pretrained='torchvision://resnet101',
backbone=dict(type='ResNet', depth=101),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=2048,
out_channels=channel_cfg['num_output_channels'],
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.0),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.2, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.0),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/atrw'
data = dict(
samples_per_gpu=64,
workers_per_gpu=2,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_train.json',
img_prefix=f'{data_root}/images/train/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_val.json',
img_prefix=f'{data_root}/images/val/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_val.json',
img_prefix=f'{data_root}/images/val/',
data_cfg=data_cfg,
pipeline=test_pipeline,
dataset_info={{_base_.dataset_info}}),
)
configs/animal/2d_kpt_sview_rgb_img/topdown_heatmap/atrw/res152_atrw_256x256.py 0 → 100755
View file @ 37c8cebc
_base_ = [
'../../../../_base_/default_runtime.py',
'../../../../_base_/datasets/atrw.py'
]
evaluation = dict(interval=10, metric='mAP', save_best='AP')
optimizer = dict(
type='Adam',
lr=5e-4,
)
optimizer_config = dict(grad_clip=None)
# learning policy
lr_config = dict(
policy='step',
warmup='linear',
warmup_iters=500,
warmup_ratio=0.001,
step=[170, 200])
total_epochs = 210
log_config = dict(
interval=1,
hooks=[
dict(type='TextLoggerHook'),
# dict(type='TensorboardLoggerHook')
])
channel_cfg = dict(
num_output_channels=15,
dataset_joints=15,
dataset_channel=[
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14],
],
inference_channel=[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14])
# model settings
model = dict(
type='TopDown',
pretrained='torchvision://resnet152',
backbone=dict(type='ResNet', depth=152),
keypoint_head=dict(
type='TopdownHeatmapSimpleHead',
in_channels=2048,
out_channels=channel_cfg['num_output_channels'],
loss_keypoint=dict(type='JointsMSELoss', use_target_weight=True)),
train_cfg=dict(),
test_cfg=dict(
flip_test=True,
post_process='default',
shift_heatmap=True,
modulate_kernel=11))
data_cfg = dict(
image_size=[256, 256],
heatmap_size=[64, 64],
num_output_channels=channel_cfg['num_output_channels'],
num_joints=channel_cfg['dataset_joints'],
dataset_channel=channel_cfg['dataset_channel'],
inference_channel=channel_cfg['inference_channel'],
soft_nms=False,
nms_thr=1.0,
oks_thr=0.9,
vis_thr=0.2,
use_gt_bbox=True,
det_bbox_thr=0.0,
bbox_file='',
)
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.0),
dict(type='TopDownRandomShiftBboxCenter', shift_factor=0.2, prob=0.3),
dict(type='TopDownRandomFlip', flip_prob=0.5),
dict(
type='TopDownHalfBodyTransform',
num_joints_half_body=8,
prob_half_body=0.3),
dict(
type='TopDownGetRandomScaleRotation', rot_factor=40, scale_factor=0.5),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(type='TopDownGenerateTarget', sigma=2),
dict(
type='Collect',
keys=['img', 'target', 'target_weight'],
meta_keys=[
'image_file', 'joints_3d', 'joints_3d_visible', 'center', 'scale',
'rotation', 'bbox_score', 'flip_pairs'
]),
]
val_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='TopDownGetBboxCenterScale', padding=1.0),
dict(type='TopDownAffine'),
dict(type='ToTensor'),
dict(
type='NormalizeTensor',
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
dict(
type='Collect',
keys=['img'],
meta_keys=[
'image_file', 'center', 'scale', 'rotation', 'bbox_score',
'flip_pairs'
]),
]
test_pipeline = val_pipeline
data_root = 'data/atrw'
data = dict(
samples_per_gpu=64,
workers_per_gpu=2,
val_dataloader=dict(samples_per_gpu=32),
test_dataloader=dict(samples_per_gpu=32),
train=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_train.json',
img_prefix=f'{data_root}/images/train/',
data_cfg=data_cfg,
pipeline=train_pipeline,
dataset_info={{_base_.dataset_info}}),
val=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_val.json',
img_prefix=f'{data_root}/images/val/',
data_cfg=data_cfg,
pipeline=val_pipeline,
dataset_info={{_base_.dataset_info}}),
test=dict(
type='AnimalATRWDataset',
ann_file=f'{data_root}/annotations/keypoint_val.json',
img_prefix=f'{data_root}/images/val/',
data_cfg=data_cfg,
pipeline=test_pipeline,
dataset_info={{_base_.dataset_info}}),
)
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment