Progressive Spatio-Temporal Graph Convolutional Network for Skeleton-Based Human Action Recognition

TL;DR

This paper introduces a progressive method to automatically discover compact, task-specific topologies for spatio-temporal graph convolutional networks, achieving high accuracy with lower computational costs in skeleton-based human action recognition.

Contribution

It presents a novel approach for automatically optimizing GCN topologies in a progressive manner, reducing complexity while maintaining or improving performance.

Findings

Competitive or superior classification accuracy

Significantly reduced computational complexity

Effective on standard skeleton-based datasets

Abstract

Graph convolutional networks (GCNs) have been very successful in skeleton-based human action recognition where the sequence of skeletons is modeled as a graph. However, most of the GCN-based methods in this area train a deep feed-forward network with a fixed topology that leads to high computational complexity and restricts their application in low computation scenarios. In this paper, we propose a method to automatically find a compact and problem-specific topology for spatio-temporal graph convolutional networks in a progressive manner. Experimental results on two widely used datasets for skeleton-based human action recognition indicate that the proposed method has competitive or even better classification performance compared to the state-of-the-art methods with much lower computational complexity.