A Survey of Human Action Recognition Based on Graph Convolutional Networks

With the development of embodied intelligence and computer vision, human action recognition has gradually become an important component of interactive cognition. Graph convolutional networks exhibit strong capabilities in handling graph data relationships, and are of great significance for research in the domain of action recognition. The paper offers an overview of action recognition using graph convo-lutional networks. Firstly, it introduces the concept of Graph Convolutional Networks. Secondly, it introduces the realm of action recognition. Afterward, it presents action recognition techniques leveraging graph convolutional networks and hy-pergraphs, and enumerates commonly used datasets for action recognition with graph convolutional networks. Finally, we present a conclusion of action recognition based on graph convolutional networks.