Joint Coordinate Regression and Association For Multi-Person Pose Estimation, A Pure Neural Network Approach
Abstract
We introduce a novel one-stage end-to-end multi-person 2D pose estimation algorithm, known as Joint Coordinate Regression and Association (JCRA), that produces human pose joints and associations without requiring any post-processing. The proposed algorithm is fast, accurate, effective, and simple. The one-stage end-to-end network architecture significantly improves the inference speed of JCRA. Meanwhile, we devised a symmetric network structure for both the encoder and decoder, which ensures high accuracy in identifying keypoints. It follows an architecture that directly outputs part positions via a transformer network, resulting in a significant improvement in performance. Extensive experiments on the MS COCO and CrowdPose benchmarks demonstrate that JCRA outperforms state-of-the-art approaches in both accuracy and efficiency. Moreover, JCRA demonstrates 69.2 mAP and is 78\% faster at inference acceleration than previous state-of-the-art bottom-up algorithms. The code for this algorithm will be publicly available.
Keywords
Cite
@article{arxiv.2307.01004,
title = {Joint Coordinate Regression and Association For Multi-Person Pose Estimation, A Pure Neural Network Approach},
author = {Dongyang Yu and Yunshi Xie and Wangpeng An and Li Zhang and Yufeng Yao},
journal= {arXiv preprint arXiv:2307.01004},
year = {2024}
}
Comments
This paper has been accepted by MMasia 2023 and is an oral presentation