mmpose/configs/animal_2d_keypoint/topdown_heatmap

Top-down heatmap-based pose estimation

Top-down methods divide the task into two stages: object detection, followed by single-object pose estimation given object bounding boxes Instead of estimating keypoint coordinates directly, the pose estimator will produce heatmaps which represent the likelihood of being a keypoint, following the paradigm introduced in Simple Baselines for Human Pose Estimation and Tracking.

Results and Models

Animal-Pose Dataset

Results on AnimalPose validation set (1117 instances)

Model	Input Size	AP	AR	Details and Download
HRNet-w32	256x256	0.740	0.780	hrnet_animalpose.md
HRNet-w48	256x256	0.738	0.778	hrnet_animalpose.md
ResNet-152	256x256	0.704	0.748	resnet_animalpose.md
ResNet-101	256x256	0.696	0.736	resnet_animalpose.md
ResNet-50	256x256	0.691	0.736	resnet_animalpose.md

AP-10K Dataset

Results on AP-10K validation set

Model	Input Size	AP	Details and Download
HRNet-w48	256x256	0.728	hrnet_ap10k.md
HRNet-w32	256x256	0.722	hrnet_ap10k.md
ResNet-101	256x256	0.681	resnet_ap10k.md
ResNet-50	256x256	0.680	resnet_ap10k.md
CSPNeXt-m	256x256	0.703	cspnext_udp_ap10k.md

Desert Locust Dataset

Results on Desert Locust test set

Model	Input Size	AUC	EPE	Details and Download
ResNet-152	160x160	0.925	1.49	resnet_locust.md
ResNet-101	160x160	0.907	2.03	resnet_locust.md
ResNet-50	160x160	0.900	2.27	resnet_locust.md

Grévy’s Zebra Dataset

Results on Grévy’s Zebra test set

Model	Input Size	AUC	EPE	Details and Download
ResNet-152	160x160	0.921	1.67	resnet_zebra.md
ResNet-101	160x160	0.915	1.83	resnet_zebra.md
ResNet-50	160x160	0.914	1.87	resnet_zebra.md