Learning Spatio-Temporal Features with 3D Residual Networks for Action Recognition

Kensho Hara, Hirokatsu Kataoka, Yutaka Satoh

URL

Proceedings: ICCV 2017 Workshop (Chalearn)

Year: 2017

Peer Reviewed : Yes

Results: 
Kinetics /Kinetics-val: 58% (3D ResNet-34)