Beyond Two-stream: Skeleton-based Three-stream Networks for Action Recognition in Videos.

Jianfeng Xu, Kazuyuki Tasaka, Hiromasa Yanagihara

Browse the full ICPR paper archive.