UIDNet in Tensorflow

The code for the AAAI-20 paper End-to-End Unpaired Image Denoising with Conditional Adversarial Networks (Aminer version).

Prerequisites

• Python 3.6
• Tensorflow 1.4.0
• Keras 2.2.4
• SciPy 1.1.0

Usage

To train a model on Low-dose CT images:

$ python main.py --ndct ./data/CUT_3mm/cut_clean/ --ldct ./data/CUT_1mm/cut_noise/ --model_name wgan-gp --c_dim 1 --epoch 50 --gpu_id 0

To test with an existing model:

$ python main.py --ndct ./data/test_1mm/cut_clean/ --ldct ./data/test_1mm/cut_noise/ --model_name wgan-gp --c_dim 1 --epoch 50 --gpu_id 0 --is_train False --checkpoint_dir ckpt_95002

You can refer to DCGAN-tensorflow for more examples.

Results of Low-dose CT image denoising

Citation

If you use this code for your research, please cite our papers.

@inproceedings{hong2020end,
  title = {End-to-End Unpaired Image Denoising with Conditional Adversarial Networks},
  author = {Zhiwei Hong and Xiaocheng Fan and Tao Jiang and Jianxing Feng,
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2020},
}