TY  - RPRT
TI  - Action Recognition in Untrimmed Videos with Composite Self-Attention Two-Stream Framework
AU  - Dong Cao
AU  - Lisha Xu
AU  - HaiBo Chen
PY  - 2019
DO  - 10.1007/978-3-030-41299-9_3
UR  - https://arxiv.org/abs/1908.04353
ID  - 1908.04353
ER  -