We introduce PyTorchVideo, an open-source deep-learning library that provides a rich set of modular, efficient, and reproducible components for a variety of video understanding tasks, including classification, detection, self-supervised learning, and low-level processing. The library covers a full stack of video understanding tools including multimodal data loading, transformations, and models that reproduce state-of-the-art performance. PyTorchVideo further supports hardware acceleration that enables real-time inference on mobile devices. The library is based on PyTorch and can be used by any training framework; for example, PyTorchLightning, PySlowFast, or Classy Vision. PyTorchVideo is available at https://pytorchvideo.org/
@article{arxiv.2111.09887,
title = {PyTorchVideo: A Deep Learning Library for Video Understanding},
author = {Haoqi Fan and Tullie Murrell and Heng Wang and Kalyan Vasudev Alwala and Yanghao Li and Yilei Li and Bo Xiong and Nikhila Ravi and Meng Li and Haichuan Yang and Jitendra Malik and Ross Girshick and Matt Feiszli and Aaron Adcock and Wan-Yen Lo and Christoph Feichtenhofer},
journal= {arXiv preprint arXiv:2111.09887},
year = {2021}
}