End-to-end Video-level Representation Learning for Action Recognition

Jiagang Zhu, Wei Zou, Zheng Zhu, Lin Li

URL

Proceedings: arxiv

Year: 2017

Peer Reviewed : No

Results: 
UCF101 /UCF101 Eval: 98% (DTPP (Kinetics pre-training))
HMDB /HMDB Eval: 82.1% (DTPP (Kinetics pre-training))