/TensorFlow-LS-InfoGAN

TensorFlow-LS-InfoGAN is a TensorFlow-based repo for generating MNIST digits using LS-GAN and InfoGAN, promoting image quality and feature disentanglement.

Primary LanguagePython

LS-GAN-TF

This repository contains some experiments of mine with Least Squares Generative Adversarial Networks (LS-GAN) with an InfoGAN extension. LS-GAN aims to generate high-quality images by minimizing the Pearson χ2 divergence, and the InfoGAN extension allows for disentangled representation learning in an unsupervised manner.

Prerequisites

Before you begin, ensure you have met the following requirements:

  • You have installed Python 3.x
  • You have installed TensorFlow 1.x (The code may need updates for TensorFlow 2.x)
  • You have a basic understanding of deep learning and GANs

Usage

To use LS-GAN-TF, follow these steps:

python Info_LSGAN_Experiment.py

You can adjust hyperparameters within the script as needed.