An attempt to upsample images by factor of 4 using GAN
Model file contains the model Generator and Discriminator code
Train file contains the code to train the model
Test file contains the code to train the model. To test, simply add your image in the input folder of the test_image folder and change the path accordingly in the code
A juypter notebook containing code is also provided for training on colab or asure, whatever you prefer
The data used for training - https://drive.google.com/drive/folders/1FZ4hENjVS-AGnlPWnbMG-YqzDRS8hKb5?usp=sharing
Link to PPT - https://drive.google.com/file/d/1Wp6utVCK8Hd2RBC49CuObuVEbTYJHmnE/view?usp=sharing
Please open PPT with google slides to view it's content properly
SRGAN is based on the Paper - https://arxiv.org/abs/1609.04802
Improvement is inspired by the Paper - https://arxiv.org/abs/1809.00219