GRA: Graph Representation Alignment for Semi-Supervised Action Recognition

doi:10.1109/TNNLS.2023.3347593

GRA: Graph Representation Alignment for Semi-Supervised Action Recognition

2024-09-01

0

PRE

AI

DOI：10.1109/TNNLS.2023.3347593

Original

Original request for help

Save

Abstract

En 中文

Graph convolutional networks (GCNs) have emerged as a powerful tool for action recognition, leveraging skeletal graphs to encapsulate human motion. Despite their efficacy, a significant challenge remains the dependency on huge labeled datasets. Acquiring such datasets is often prohibitive, and the frequent occurrence of incomplete skeleton data, typified by absent joints and frames, complicates the testing phase. To tackle these issues, we present graph representation alignment (GRA), a novel approach with two main contributions: 1) a self-training (ST) paradigm that substantially reduces the need for labeled data by generating high-quality pseudo-labels, ensuring model stability even with minimal labeled inputs and 2) a representation alignment (RA) technique that utilizes consistency regularization to effectively reduce the impact of missing data components. Our extensive evaluations on the NTU RGB+D and Northwestern-UCLA (N-UCLA) benchmarks demonstrate that GRA not only improves GCN performance in data-constrained environments but also retains impressive performance in the face of data incompleteness.

Keywords:

Action recognition

consistency regularization

graph convolutional networks (GCNs)

graph representation learning

self-training (ST)

semi-supervised learning

skeleton action recognition

GRA: Graph Representation Alignment for Semi-Supervised Action Recognition

Abstract

Journal

Organization