EPIT

October 6, 2023 ยท View on GitHub

This is the official implementation of "Learning Non-Local Spatial-Angular Correlation for Light Field Image Super-Resolution", ICCV 2023.

[paper] [arXiv] [project]

Training & Evaluation

  • Download the EPFL, HCInew, HCIold, INRIA and STFgantry datasets via Baidu Drive (key:7nzy) or OneDrive, and place the 5 datasets to the folder ./datasets/.
  • Run Generate_Data_for_SSR_Training.py to generate training data, and begin to train the EPIT (on 5x5 by default) for 2x/4x SR:
  $ python train.py --scale_factor \$2/4$
  • Run Generate_Data_for_SSR_Test.py to generate evaluation data, and you can quick run test.py to perform network inference by using our released models.
  python test.py

Quantitative Results


Visual Comparison


Citiation

If you find this work helpful, please consider citing:

@InProceedings{Liang_2023_ICCV,
    author    = {Liang, Zhengyu and Wang, Yingqian and Wang, Longguang and Yang, Jungang and Zhou, Shilin and Guo, Yulan},
    title     = {Learning Non-Local Spatial-Angular Correlation for Light Field Image Super-Resolution},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
    month     = {October},
    year      = {2023},
    pages     = {12376-12386}
}

Contact

Welcome to raise issues or email to zyliang@nudt.edu.cn for any questions regarding our EPIT.