A Spherical Convolution Approach for Learning Long Term Viewport Prediction in 360 Immersive Video

Wu, Chenglei, et al. "A Spherical Convolution Approach for Learning Long Term Viewport Prediction in 360 Immersive Video." Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 34. No. 01. 2020.

1. Requirements

   • pytorch
   • s2cnn

2. Training and Testing

As this is my first AI project, so many of the old training and testing codes are easy to use and require refactoring. So this repo only contains the some of essential models. Please refer to this new project for cleaner codes and many other improvements.

3. Citation If you find these codes useful, please cite:

@inproceedings{wu2020spherical,
  title={A Spherical Convolution Approach for Learning Long Term Viewport Prediction in 360 Immersive Video},
  author={Wu, Chenglei and Zhang, Ruixiao and Wang, Zhi and Sun, Lifeng},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={34},
  number={01},
  pages={14003--14040},
  year={2020}
}

4. Contact

If you have any questions, feel free to contact me via wucl18@mails.tsinghua.edu.cn.