Interweaved Graph and Attention Network for 3D Human Pose Estimation

TL;DR

This paper introduces IGANet, a novel network combining graph convolutional networks and attention mechanisms to improve 3D human pose estimation by capturing both local and global correlations, achieving state-of-the-art results.

Contribution

The paper proposes a new interweaved graph and attention network with an IGA module and uMLP for enhanced skeleton representation learning in 3D pose estimation.

Findings

Achieves state-of-the-art performance on Human3.6M dataset.

Outperforms previous methods on MPI-INF-3DHP dataset.

Effectively captures multi-granularity joint information.

Abstract

Despite substantial progress in 3D human pose estimation from a single-view image, prior works rarely explore global and local correlations, leading to insufficient learning of human skeleton representations. To address this issue, we propose a novel Interweaved Graph and Attention Network (IGANet) that allows bidirectional communications between graph convolutional networks (GCNs) and attentions. Specifically, we introduce an IGA module, where attentions are provided with local information from GCNs and GCNs are injected with global information from attentions. Additionally, we design a simple yet effective U-shaped multi-layer perceptron (uMLP), which can capture multi-granularity information for body joints. Extensive experiments on two popular benchmark datasets (i.e. Human3.6M and MPI-INF-3DHP) are conducted to evaluate our proposed method.The results show that IGANet achieves state-of-the-art performance on both datasets. Code is available at https://github.com/xiu-cs/IGANet.