PyTorch implemented C3D, R3D, R2Plus1D models for video activity recognition. - View it on GitHub