This course project entails a comparitive study of how the results vary when discriminator sees differently pre-processed images from the dataset(s). These pre-processings will be low-level digital image processing algorithms which this course is centered around.
- GAN
- DCGAN
- WGAN
- WGAN-GP
- SRGAN
- infoGAN
- RealnessGAN
- Cycle-GAN + Style-GAN
- pix2pix
- MNIST
- Fashion-MNIST
- AFHQ (Dog Subset)
- Facades
- SR-Kaggle
PyTorch Lightning + WandB
++ Literature Survey/Review
Aryan Garg
B19153