An Improved Graph Pooling Network for Skeleton-Based Action Recognition

TL;DR

This paper introduces IGPN, an improved graph pooling network for skeleton-based action recognition that enhances feature processing through structural partitioning, adaptive weighting, and information fusion, leading to significant accuracy gains and efficiency.

Contribution

The paper presents a novel graph pooling method with region-awareness, adaptive weighting, and information fusion modules, seamlessly integrated with existing GCN models for better skeleton-based action recognition.

Findings

Achieves higher accuracy on NTU-RGB+D 60 dataset

Reduces computational Flops by nearly 70%

Enhances feature discrimination with new modules

Abstract

Pooling is a crucial operation in computer vision, yet the unique structure of skeletons hinders the application of existing pooling strategies to skeleton graph modelling. In this paper, we propose an Improved Graph Pooling Network, referred to as IGPN. The main innovations include: Our method incorporates a region-awareness pooling strategy based on structural partitioning. The correlation matrix of the original feature is used to adaptively adjust the weight of information in different regions of the newly generated features, resulting in more flexible and effective processing. To prevent the irreversible loss of discriminative information, we propose a cross fusion module and an information supplement module to provide block-level and input-level information respectively. As a plug-and-play structure, the proposed operation can be seamlessly combined with existing GCN-based models. We conducted extensive evaluations on several challenging benchmarks, and the experimental results indicate the effectiveness of our proposed solutions. For example, in the cross-subject evaluation of the NTU-RGB+D 60 dataset, IGPN achieves a significant improvement in accuracy compared to the baseline while reducing Flops by nearly 70%; a heavier version has also been introduced to further boost accuracy.