Deep-Reinforcement-Learning-for-Image-Denoising-via-Residual-Recovery (R3L) in pytorch.

1. This is a simple pytorch implementation of DRL (PPO is used) for image denoising via residual recovery.
2. Detailed illustration can be found in our paper R3L: Connecting Deep Reinforcement Learning To Recurrent Neural Networks For Image Denoising Via Residual Recovery (accepted by ICIP 2021).
3. Although this project is for a specific task, this framework is designed ASAP (as simple as possible) to be applied for different tasks trained in "Batch Environment" (Batch * Channel * Height * Width) by slightly modifing the corresponding network and envrionment.

Introduction to this implementation:

1. Current implementations of PPO usually focus on environments with states in shape of (Height * Width) raising a gap for implementations in CV where (Channel * Height * Width) is needed.
2. This implementation aims for an easy-to-modify PPO framework for CV tasks.
3. The PPO used here is PPO-clip.

How to apply to other tasks:

1. Customize the environment by setting task specific reset(), step() in environment.py.
2. Customize the data file paths in PPO_batch.py.
3. Customize data argumentation in Load_batch.py.

Dependance:

1. pytorch >= 1.6
2. opencv

Citation

In case of use, please cite our publication:

Rongkai Zhang, Jiang Zhu, Zhiyuan Zha, Justin Dauwels, Bihan Wen, "R3L: Connecting Deep Reinforcement Learning to Recurrent Neural Networks for Image Denoising via Residual Recovery," ICIP 2021.

Bibtex:

@inproceedings{zhang2021r3l,
  title={R3L: Connecting Deep Reinforcement Learning to Recurrent Neural Networks for Image Denoising via Residual Recovery},
  author={Zhang, Rongkai and Zhu, Jiang and Zha, Zhiyuan and Dauwels, Justin and Wen, Bihan},
  booktitle={2021 IEEE International Conference on Image Processing (ICIP)},
  pages={1624--1628},
  year={2021},
  organization={IEEE}
}