Liqun Chen, Yuyao Hu, Jiewen Nie, Tianfan Xue and Jinwei Gu

The codes was tested on Windows 10, with Python and PyTorch. Packages required to reproduce the results can be found in requirements.txt. The following software / hardware is tested and recommended:

• numpy
• tqdm
• Python >= 3.9
• matplotlib >= 3.5
• pytorch >= 2.0
• torchvision >= 0.15
• pandas >= 1.4

This repository contains codes for LWNet.

LWNet
|   README.md
|   requirements.txt
|   main_lwnet.py
|   statistic_zer_rule.py
|   demo.gif
|---data  
|---model
|---results  

/data include input data (GT).

/models include models for Stage_I and Stage_II.

/results store the optimization results.

To test LWNet and reproduce some results shown in the paper:

• Run main_lwnet.py. The outputs will be saved in results folders
• Modify parameters in configs/lwnet.yaml, run main_lwnet.py. The outputs will be saved in results folders

For any question, you can contact chenliqun@pjlab.org.cn

If you use this codebase or any part of it for a publication, please cite:

@article{chen2024learning,
  title={Learning-based lens wavefront aberration recovery},
  author={Chen, Liqun and Hu, Yuyao and Nie, Jiewen and Xue, Tianfan and Gu, Jinwei},
  journal={Optics Express},
  volume={32},
  number={11},
  pages={18931--18943},
  year={2024},
  publisher={Optica Publishing Group}
}