Reconstructing compressed CUFS images using a Variational Autoencoder (VAE).
Used PyTorch framework for development and Google Colaboratory to use GPU while training the model.
Please note:
- Before running the Notebook, download the CUFS dataset images (http://mmlab.ie.cuhk.edu.hk/archive/facesketch.html) and put them into the 'imgs' subdirectory at 'CIS680/HW3/cufs' and upload the CIS680 directory to Google Drive.