Part-aligned pose-guided recurrent network for action recognition.
Pattern Recognition(2019)
摘要
•The novel end-to-end architecture can improve the accuracy of action recognition efficiently.•Introducing the part alignment into action recognition can capture spatio-temporal evolutions of actions.•The part-based hierarchical pooling approach can learn a robust and discriminative feature.•Our method obtains the state-of-the-art results on two important benchmarks of action recognition.
更多查看译文
关键词
Action recognition,Part alignment,Auto-transformer attention
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络