Hybrid Representation Learning for End-to-End Multi-Person Pose Estimation
IEEE Transactions on Circuits and Systems for Video Technology · 2025-07 · 35 (7) · 6437-6451
DOI: 10.1109/tcsvt.2025.3536381 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Circuits and Systems for Video Technology, 2025-07. Verfasst von Qiyuan Dai und 1 weiteren Autor:innen. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 75 Literaturangaben. Gefördert von: Key Science and Technology Program of Anhui, National Natural Science Foundation of China, USTC-JAC Smart Electric Vehicle Joint Laboratory.
Abstract (Originalsprache)
Recent multi-person pose estimation methods design end-to-end pipelines under the DETR framework. However, these methods involve complex keypoint decoding processes because the DETR framework cannot be directly used for pose estimation, which results in constrained performance and ineffective information interaction between human instances. To tackle this issue, we propose a hybrid representation learning method for end-to-end multi-person pose estimation. Our method represents instance-level and keypoint-level information as hybrid queries based on point set prediction and can facilitate parallel interaction between instance-level and keypoint-level representations in a unified decoder. We also employ the instance segmentation task for auxiliary training to enrich the spatial context of hybrid representations. Furthermore, we introduce a pose-unified query selection (PUQS) strategy and an instance-gated module (IGM) to improve the keypoint decoding process. PUQS predicts local pose proposals to produce scale-aware instance initializations and can avoid the scale assignment mistake of one-to-one matching. IGM refines instance contents and filters out invalid information with the message of cross-instance interaction and can enhance the decoder’s capability to handle queries of instances. Compared with current end-to-end multi-person pose estimation methods, our method can detect human instances and body keypoints simultaneously through a concise decoding process. Extensive experiments on COCO Keypoint and CrowdPose benchmarks demonstrate that our method outperforms some state-of-the-art methods.
Förderung
- Key Science and Technology Program of Anhui
- National Natural Science Foundation of China
- USTC-JAC Smart Electric Vehicle Joint Laboratory
Zitieren
APA
Dai, Q., & Ling, Q. (2025). Hybrid Representation Learning for End-to-End Multi-Person Pose Estimation. IEEE Transactions on Circuits and Systems for Video Technology, 35(7), 6437-6451. https://doi.org/10.1109/tcsvt.2025.3536381
Vancouver
Dai Q, Ling Q. Hybrid Representation Learning for End-to-End Multi-Person Pose Estimation. IEEE Transactions on Circuits and Systems for Video Technology. 2025;35(7):6437-6451. doi:10.1109/tcsvt.2025.3536381
BibTeX
@article{Dai2025,
author = {Dai, Qiyuan and Ling, Qiang},
title = {Hybrid Representation Learning for End-to-End Multi-Person Pose Estimation},
journal = {IEEE Transactions on Circuits and Systems for Video Technology},
year = {2025},
volume = {35},
number = {7},
pages = {6437--6451},
doi = {10.1109/tcsvt.2025.3536381}
}
EUVIMED