期刊
IEEE SIGNAL PROCESSING LETTERS
卷 28, 期 -, 页码 2058-2062出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/LSP.2021.3116513
关键词
Focusing; Convolution; Skeleton; Transformers; Hidden Markov models; Context modeling; Aggregates; Focusing and diffusion; bidirectional attention; graph convolutional network; action recognition
资金
- National Natural Science Foundation of China [61772513]
This letter proposes a FDGCN to address the spatial-temporal context exploration issue in skeleton-based action recognition, which processes each skeleton frame through focusing and diffusion processes, combines with a Transformer encoder layer to capture temporal context, and ultimately achieves context transfer among spatial joints.
Graph Convolutional Networks have been successfully applied in skeleton-based action recognition. The key is fully exploring the spatial-temporal context. This letter proposes a Focusing-Diffusion Graph Convolutional Network (FDGCN) to address this issue. Each skeleton frame is first decomposed into two opposite-direction graphs for subsequent focusing and diffusion processes. Next, the focusing process generates a spatial-level representation for each frame individually by an attention module. This representation is regarded as a supernode to aggregate the feature from each joint node in each frame for spatial context extraction. After generating supernodes for the entire sequence, a transformer encoder layer is proposed to capture the temporal context further. Finally, these supernodes pass the embedded spatial-temporal context back to the spatial joints through the diffusion graph in the diffusing process. Extensive experiments on the NTU RGB+D and Skeleton-Kinetics benchmarks demonstrate the effectiveness of our approach.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据