AOD-GAN is based on AOD-Net:An All-in-One Network for Dehazing and Beyond. AOD-Net is a lightly model application to image dehaze.we extend the model and called AOD-GAN.
Compare
| model | Architecture | GAN Loss | MSE Loss | Perception Loss |
|---|---|---|---|---|
| AOD-GAN | Only G | NO | Yes | NO |
| AOD-Net | G + D | Yes | Yes | Yes |
Environmention:
- Python 3
- Pytorch 0.4
- ubuntu 16.04
Test On Image
python TestImage.py --input input_image --output_image
Test On Video
python TestVideo.py --input input_video.mp4
Results
References
- An All-in-One Network for Dehazing and Beyond. Boyi Li, Xiulian Peng, Zhangyang Wang, Jizheng Xu, Dan Feng.
- Generative Adversarial Networks. Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, Yoshua Bengio.
- YOLOv3:An Incremental Improvement. Joseph Redmon, Ali Farhadi.