Skeleton-based action recognition with local dynamic spatial-temporal aggregation

Open Webpage

Lianyu Hu, Shenglan Liu, Wei Feng

Published: 2023, Last Modified: 08 Apr 2025Expert Syst. Appl. 2023EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Previous spatial–temporal convolutional networks suffer from fixed graph strategy.•Naive spatial–temporal graph attention networks does not work well.•Strategies to release the potential of spatial–temporal graph attention networks.•Much progress on classifying similar action pairs.