3D human pose estimation in motion based on multi-stage regression |
| |
| Affiliation: | 1. Department of Information Engineering, Shangqiu Institute of Technology, Shangqiu 476000, China;2. Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China;3. YouSan Educational Technology, Beijing 100080, China;1. Institute of Logistics Science and Engineering, Shanghai Maritime University, Shanghai 201306, China;2. Anhui Longquan Silicon Materials Co.,Ltd, Huaiyuan, Anhui 233400, China;1. School of Astronautics, Harbin Institute of Technology, 92 Xidazhi Street, Harbin 150080, China;2. SHENZHEN LAIBAO Hi-TECH Corporation, Xili Street, Nanshan District, Shenzhen 518000, China;1. Artificial Intelligence Institute, Shanghai Jiao Tong University, Shanghai 200240, China;2. Shenzhen Eye Hospital, The Second Affiliated Hospital of Jinan University, Shenzhen 518040, China |
| |
| Abstract: | 3D human pose estimation in motion is a hot research direction in the field of computer vision. However, the performance of the algorithm is affected by the complexity of 3D spatial information, self-occlusion of human body, mapping uncertainty and other problems. In this paper, we propose a 3D human joint localization method based on multi-stage regression depth network and 2D to 3D point mapping algorithm. First of all, we use a single RGB image as the input, through the introduction of heatmap and multi-stage regression to constantly optimize the coordinates of human joint points. Then we input the 2D joint points into the mapping network for calculation, and get the coordinates of 3D human body joint points, and then to complete the 3D human body pose estimation task. The MPJPE of the algorithm in Human3.6 M dataset is 40.7. The evaluation of dataset shows that our method has obvious advantages. |
| |
| Keywords: | 3D human pose estimation Multi-stage regression Heatmap 2D to 3D mapping |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
