In this repository, I have implemented variational autoencoder(convolutional) to generate desired images from their respective given images
Steps involved in data preprocessing:-
you can build the dataset organized in a way such that:-
1. input image folder
2. ground truth image folder
Normalize the data
Split the data into train and validation
Dataset is private so i cannot give more information about that.
That's it.
Go through the notebook for more details and implementation.