Conditional Directed Graph Convolution for 3D Human Pose Estimation

TL;DR

This paper introduces a novel directed graph convolutional network that explicitly models the hierarchical structure of human skeletons for improved 3D pose estimation from monocular videos, achieving top performance on benchmark datasets.

Contribution

It proposes a directed graph representation of the human skeleton and a spatial-temporal conditional graph convolution to better capture pose dependencies and hierarchy.

Findings

Outperforms existing methods on Human3.6M and MPI-INF-3DHP datasets.

Directed graphs better exploit skeletal hierarchy than undirected graphs.

Conditional graph topology adapts to different poses, improving accuracy.

Abstract

Graph convolutional networks have significantly improved 3D human pose estimation by representing the human skeleton as an undirected graph. However, this representation fails to reflect the articulated characteristic of human skeletons as the hierarchical orders among the joints are not explicitly presented. In this paper, we propose to represent the human skeleton as a directed graph with the joints as nodes and bones as edges that are directed from parent joints to child joints. By so doing, the directions of edges can explicitly reflect the hierarchical relationships among the nodes. Based on this representation, we further propose a spatial-temporal conditional directed graph convolution to leverage varying non-local dependence for different poses by conditioning the graph topology on input poses. Altogether, we form a U-shaped network, named U-shaped Conditional Directed Graph Convolutional Network, for 3D human pose estimation from monocular videos. To evaluate the effectiveness of our method, we conducted extensive experiments on two challenging large-scale benchmarks: Human3.6M and MPI-INF-3DHP. Both quantitative and qualitative results show that our method achieves top performance. Also, ablation studies show that directed graphs can better exploit the hierarchy of articulated human skeletons than undirected graphs, and the conditional connections can yield adaptive graph topologies for different poses.