The Root Element of Human Poses is Radian: MCPRL is All You Need

Ziming Cheng, Xiangning Ruan, Qixiang Yin, Zhicheng ZhaoICME 2024 · OpenReview

Abstract

3D Human Pose Estimation aims to determine the spatial coordinates of key anatomical landmarks on human body. Common benchmarks for this task include Human3.6M and MPI-INF-3DHP, while substantial redundancy exists. Additionally, both datasets are confined indoors due to equipment limitations, compromising diversity and generalization. To address the above issues, we first quantify dataset redundancy by introducing Generalized Radian Pruning (GRP), which employs a novel radians-based method to categorize and optimize human poses. Secondly, based on merging H36M and MPII datasets, we construct a new Manifold Cadre Poses with Radian List-systematically (MCPRL) dataset where missing outdoor scenarios, particularly dynamic collision actions are supplemented. Five strong baselines are compared and the experimental results demonstrate the effectiveness of the GRP method and the superiority of the MCPRL dataset. Our released dataset reduces data redundancy by 35%, surpasses H36M and MPII by 114% in diversity, with an average 11.19mm reduction in the MPJPE indicator. MCPRL dataset can be found at https://github.com/Rxn666/MCPRL.
BibTeX
@inproceedings{cheng2024root,
  author = {Ziming Cheng and Xiangning Ruan and Qixiang Yin and Zhicheng Zhao},
  title = {{The Root Element of Human Poses is Radian: MCPRL is All You Need}},
  booktitle = {ICME},
  year = {2024},
  url = {https://openreview.net/forum?id=SowYxjaEKB},
}

Loading related content...

Figure