SlowFast Networks for Video Recognition
This addresses the problem of efficient and accurate video recognition for applications in action classification and detection, with incremental improvements in model design.
The paper tackles video recognition by proposing SlowFast networks, which use a slow pathway for spatial semantics and a fast pathway for motion, achieving state-of-the-art accuracy on benchmarks like Kinetics, Charades, and AVA.
We present SlowFast networks for video recognition. Our model involves (i) a Slow pathway, operating at low frame rate, to capture spatial semantics, and (ii) a Fast pathway, operating at high frame rate, to capture motion at fine temporal resolution. The Fast pathway can be made very lightweight by reducing its channel capacity, yet can learn useful temporal information for video recognition. Our models achieve strong performance for both action classification and detection in video, and large improvements are pin-pointed as contributions by our SlowFast concept. We report state-of-the-art accuracy on major video recognition benchmarks, Kinetics, Charades and AVA. Code has been made available at: https://github.com/facebookresearch/SlowFast