Official tensorflow implementation codes for Pattern Recognition Letters accepted paper "Customization of latent space in semi-supervised variational autoencoder"

Simple VAE implementation in tensorflow, jax and pytorch where both the encoder and decoder model use gaussian distributions.

This project involves an exploratory analysis of the capabilities of Variational Autoencoders (VAEs) and Wasserstein Generative Adversarial Networks (WGANs) for image generation. The CelebA dataset is utilized for training and evaluating the models. The aim is to understand the strengths and weaknesses of VAEs and WGANs.