PyTorch implementation of "SlowFast Networks for Video Recognition".
- Dataset should be orgnized as:
dataset(e.g. UCF-101)
│ │ train/training
│ │ │ ApplyEyeMakeup
│ │ │ ApplyLipstick
│ │ │ ...
│ │ validation
│ │ ApplyEyeMakeup
│ │ ApplyLipstick
│ │ ...
- Modify the params in config.py and
mode
oftrain_dataloader
orval_dataloader
in train.py.
python 3
PyTorch >= 0.4.1
tensorboardX
OpenCV
[1] https://github.com/Guocode/SlowFast-Networks/
[2] https://github.com/jfzhang95/pytorch-video-recognition
[3] https://github.com/irhumshafkat/R2Plus1D-PyTorch