Papers › TransPose: Keypoint Localization via Transformer

TransPose: Keypoint Localization via Transformer

28 Dec 2020ICCV 2021 10arXiv:2012.14214archive 2025-07-28

Sen yang, Zhibin Quan, Mu Nie, Wankou Yang

While CNN-based models have made remarkable progress on human pose estimation, what spatial dependencies they capture to localize keypoints remains unclear. In this work, we propose a model called \textbf{TransPose}, which introduces Transformer for human pose estimation. The attention layers built in Transformer enable our model to capture long-range relationships efficiently and also can reveal what dependencies the predicted keypoints rely on. To predict keypoint heatmaps, the last attention layer acts as an aggregator, which collects contributions from image clues and forms maximum positions of keypoints. Such a heatmap-based localization approach via Transformer conforms to the principle of Activation Maximization~\cite{erhan2009visualizing}. And the revealed dependencies are image-specific and fine-grained, which also can provide evidence of how the model handles special cases, e.g., occlusion. The experiments show that TransPose achieves 75.8 AP and 75.0 AP on COCO validation and test-dev sets, while being more lightweight and faster than mainstream CNN architectures. The TransPose model also transfers very well on MPII benchmark, achieving superior performance on the test set when fine-tuned with small training costs. Code and pre-trained models are publicly available\footnote{\url{https://github.com/yangsenius/TransPose}}.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

yangsenius/TransPose officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Keypoint DetectionMulti-Person Pose EstimationPose Estimation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Keypoint Detection COCO (Common Objects in Context) TransPose(256x192) Test AP 75.0 #10 of 24 Archive leaderboard report
Keypoint Detection COCO (Common Objects in Context) TransPose(256x192) Validation AP 75.8 #10 of 24 Archive leaderboard report
Multi-Person Pose Estimation CrowdPose TransPose-H AP Easy 79.5 #10 of 28 Archive leaderboard report
Multi-Person Pose Estimation CrowdPose TransPose-H AP Hard 62.2 #10 of 28 Archive leaderboard report
Multi-Person Pose Estimation CrowdPose TransPose-H AP Medium 72.9 #10 of 28 Archive leaderboard report
Multi-Person Pose Estimation CrowdPose TransPose-H mAP @0.5:0.95 71.8 #10 of 28 Archive leaderboard report
Multi-Person Pose Estimation OCHuman TransPose-H AP50 82.7 #3 of 8 Archive leaderboard report
Multi-Person Pose Estimation OCHuman TransPose-H AP75 67.1 #3 of 8 Archive leaderboard report
Pose Estimation COCO test-dev TransPose-H-A6 AP 75 #21 of 47 Archive leaderboard report
Pose Estimation COCO test-dev TransPose-H-A6 AP50 92.2 #21 of 47 Archive leaderboard report
Pose Estimation COCO test-dev TransPose-H-A6 AP75 82.3 #21 of 47 Archive leaderboard report
Pose Estimation COCO test-dev TransPose-H-A6 APL 81.1 #21 of 47 Archive leaderboard report
Pose Estimation COCO test-dev TransPose-H-A6 APM 71.3 #21 of 47 Archive leaderboard report
Pose Estimation MPII Human Pose TransPose PCKh-0.5 93.5 #5 of 46 Archive leaderboard report
Pose Estimation OCHuman TransPose-H Validation AP 62.3 #19 of 19 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections