The major goal for this project is to experimenting various Variational Auto Encoder structure in the purpose of repairing images. We want to find a robust way on repairing images because image restoration has high application value in medical imaging, movie production and many other areas.
The code involves the implemntation of VAE, VAE-C and Fisher-Auto Encoders. The models are evaluated according to their ability on recontrusting damaged images from KMNIST dataset.
- Sijie Chen ([email protected]) @SijieChen
- Haoyu Jiang ([email protected]) @HaoyuJiang
This project is licensed under the [MIT + file LICENSE] License - see the LICENSE.md file for details