CGFusionFormer: Exploring Compact Spatial Representation for Robust 3D Human Pose Estimation with Low Computation Complexity

Transformer-based 2D-to-3D lifting methods have demonstrated outstanding performance in 3D human pose estimation from 2D pose sequences. However, they still encounter challenges with the relatively poor quality of 2D joints and substantial computational costs. In this paper, we propose a CGFusionFor...

Celý popis