【论文笔记】CVPR2020 Skeleton-Based Action Recognition with Shift Graph Convolutional Network

用图卷积做action recogntion

Contribution

• 提出了shift-GCN（spatial shift graph operations & temporal shift graph operations）that exceeds the state-of-the-art methods with more than 10× less computational cost，主要解决了过去工作的两个问题
  1. heavy computational complexity of GCN-based methods
  2. inflexible and predefined receptive fields of both spatial graph and temporal graph

Methods

作者灵感来源是shift-CNNs，分别就 spatial GCN 和 temporal GCN 提出新的想法，各自提了2个：

• two kinds of spatial shift graph operation
  

  1. local shift graph operation
     让physically-connected nodes 之间进行信息shift，然后再用1x1conv 来进行channel与channel之间的信息融合和升维降维。