DGFormer: Dynamic graph transformer for 3D human pose estimation

Zhangmeng Chen, Ju Dai, Junxuan Bai, Junjun Pan

Abstract: Highlights•We propose a novel dynamic Graph Transformer model for 3D human pose estimation (3D-HPE).•The Transformer is applied to exploit the global relationships among skeleton joints.•The proposed immobile GCN captures the local physical connections.•The proposed dynamic GCN learns the sparse dynamic K-nearest neighbor interactions.•Our method outperforms state-of-the-art methods for image-based 3D-HPE.