In this projet, we focused on the implementation of two popular deep learning models of DnCNN and GANs.
DnCNN is a deep learning model which is especially designed for denoising images. It uses Squeeze and Exapand Blocks, along with Residual Block
The GAN model consists of a Generator and a Discriminator, where the Generator is optimised to give as small of a MSSIM loss as possible
Team members:
- Harsh Vardhan Gupta
- Wai Kit Lam
- Ming Chak Ho
