Channel-wise Topology Refinement Graph Convolution for Skeleton-Based Action Recognition

TL;DR

This paper introduces CTR-GC, a novel graph convolution method that dynamically refines channel-wise topologies for skeleton-based action recognition, significantly improving performance over existing methods.

Contribution

The paper proposes a new CTR-GC method that models channel-specific topologies with minimal extra parameters, enhancing feature aggregation in GCNs for action recognition.

Findings

Outperforms state-of-the-art on NTU RGB+D datasets

Reduces complexity of modeling channel-wise topologies

Enhances representation capability of graph convolutions

Abstract

Graph convolutional networks (GCNs) have been widely used and achieved remarkable results in skeleton-based action recognition. In GCNs, graph topology dominates feature aggregation and therefore is the key to extracting representative features. In this work, we propose a novel Channel-wise Topology Refinement Graph Convolution (CTR-GC) to dynamically learn different topologies and effectively aggregate joint features in different channels for skeleton-based action recognition. The proposed CTR-GC models channel-wise topologies through learning a shared topology as a generic prior for all channels and refining it with channel-specific correlations for each channel. Our refinement method introduces few extra parameters and significantly reduces the difficulty of modeling channel-wise topologies. Furthermore, via reformulating graph convolutions into a unified form, we find that CTR-GC relaxes strict constraints of graph convolutions, leading to stronger representation capability. Combining CTR-GC with temporal modeling modules, we develop a powerful graph convolutional network named CTR-GCN which notably outperforms state-of-the-art methods on the NTU RGB+D, NTU RGB+D 120, and NW-UCLA datasets.