Pose Estimation on AIC

30개 결과 · ⬇ CSV · JSON

Rank	Model	AP	AP75	AR	AR50	AP50	Paper	Code	Year
1	Hulk(Finetune, ViT-L)	37.1	–	–	–	–	Hulk: A Universal Knowledge Translator for Human-Centric Tasks	opengvlab/humanbench · opengvlab/hulk	2023
2	Hulk(Finetune, ViT-B)	35.6	–	–	–	–	Hulk: A Universal Knowledge Translator for Human-Centric Tasks	opengvlab/humanbench · opengvlab/hulk	2023
3	HRFormer (HRFomer-B)	34.4	24.8	38.7	80.9	78.3	HRFormer: High-Resolution Transformer for Dense Prediction	HRNet/HRFormer	2021
4	UniHCP (finetune)	33.6	–	–	–	–	UniHCP: A Unified Model for Human-Centric Perceptions	opengvlab/unihcp	2023
5	HRNet (HRNet-w48 )	33.5	23.6	37.9	80.0	78.0	Deep High-Resolution Representation Learning for Human Pose Estimation	open-mmlab/mmdetection · PaddlePaddle/PaddleDetection · open-mmlab/mmpose · +36	2019
6	HRNet (HRNet-w32)	32.3	21.9	36.6	78.9	76.2	Deep High-Resolution Representation Learning for Human Pose Estimation	open-mmlab/mmdetection · PaddlePaddle/PaddleDetection · open-mmlab/mmpose · +36	2019
7	HRFormer (HRFomer-S)	31.6	20.9	35.8	78.0	–	HRFormer: High-Resolution Transformer for Dense Prediction	HRNet/HRFormer	2021
8	SimpleBaseline (ResNet-152)	29.9	18.3	34.3	76.9	73.8	Simple Baselines for Human Pose Estimation and Tracking	PaddlePaddle/PaddleDetection · open-mmlab/mmpose · osmr/imgclsmob · +24	2018
9	SimpleBaseline (ResNet-101)	29.4	17.4	33.7	76.3	73.6	Simple Baselines for Human Pose Estimation and Tracking	PaddlePaddle/PaddleDetection · open-mmlab/mmpose · osmr/imgclsmob · +24	2018
10	SimpleBaseline (ResNet-50)	28.0	15.8	32.1	74.1	71.6	Simple Baselines for Human Pose Estimation and Tracking	PaddlePaddle/PaddleDetection · open-mmlab/mmpose · osmr/imgclsmob · +24	2018
11	Hulk(Finetune, ViT-L)	37.1	–	–	–	–	Hulk: A Universal Knowledge Translator for Human-Centric Tasks	opengvlab/humanbench · opengvlab/hulk	2023
12	Hulk(Finetune, ViT-B)	35.6	–	–	–	–	Hulk: A Universal Knowledge Translator for Human-Centric Tasks	opengvlab/humanbench · opengvlab/hulk	2023
13	HRFormer (HRFomer-B)	34.4	24.8	38.7	80.9	78.3	HRFormer: High-Resolution Transformer for Dense Prediction	HRNet/HRFormer	2021
14	UniHCP (finetune)	33.6	–	–	–	–	UniHCP: A Unified Model for Human-Centric Perceptions	opengvlab/unihcp	2023
15	HRNet (HRNet-w48 )	33.5	23.6	37.9	80.0	78.0	Deep High-Resolution Representation Learning for Human Pose Estimation	open-mmlab/mmdetection · PaddlePaddle/PaddleDetection · open-mmlab/mmpose · +36	2019
16	HRNet (HRNet-w32)	32.3	21.9	36.6	78.9	76.2	Deep High-Resolution Representation Learning for Human Pose Estimation	open-mmlab/mmdetection · PaddlePaddle/PaddleDetection · open-mmlab/mmpose · +36	2019
17	HRFormer (HRFomer-S)	31.6	20.9	35.8	78.0	–	HRFormer: High-Resolution Transformer for Dense Prediction	HRNet/HRFormer	2021
18	SimpleBaseline (ResNet-152)	29.9	18.3	34.3	76.9	73.8	Simple Baselines for Human Pose Estimation and Tracking	PaddlePaddle/PaddleDetection · open-mmlab/mmpose · osmr/imgclsmob · +24	2018
19	SimpleBaseline (ResNet-101)	29.4	17.4	33.7	76.3	73.6	Simple Baselines for Human Pose Estimation and Tracking	PaddlePaddle/PaddleDetection · open-mmlab/mmpose · osmr/imgclsmob · +24	2018
20	SimpleBaseline (ResNet-50)	28.0	15.8	32.1	74.1	71.6	Simple Baselines for Human Pose Estimation and Tracking	PaddlePaddle/PaddleDetection · open-mmlab/mmpose · osmr/imgclsmob · +24	2018