Accepted Paper: An Articulated Structure-Aware Network for 3D Human Pose Estimation

Back to list of accepted papers

• Session 2: Multi-task Learning, NLP, Computer Vision, Applications -- Day 2 (Nov.18), talks: 09:00-11:00 (5th floor Hall 2), poster session: 11:00-13:30
• Poster number: Mon36
• Download paper

Authors

Zhenhua Tang (College of Computer Science and Software Engineering, Shenzhen University); Xiaoyan Zhang (College of Computer Science and Software Engineering, Shenzhen University); Junhui Hou (City University of Hong Kong, Hong Kong)

Abstract

In this paper, we propose a new end-to-endarticulated structure-aware network to regress 3D joint coordinates from the given 2D joint detections. The proposed method is capable of dealing with hard joints well that usually fail existing methods. Specifically, our framework cascades a refinement network with a basic network for two types of joints, and employs a attention module to simulate a camera projection model. In addition, we propose to use a random enhancement module to intensify the constraints between joints. Experimental results on the Human3.6M and HumanEva databases demonstrate the effectiveness and flexibility of the proposed network, and errors of hard joints and bone lengths are significantly reduced, compared with state-of-the-art approaches.