Simple Conditional_GAN

Tensorflow implementation of Conditional Generative Adversarial Network(CGAN).

This repository is minor change of Vanilla_GAN.
This model can generate MNIST samples with certian labels.

File discription

• main.py: Main function of implemenation, construct and train the model, generates images
• model.py: CGAN class
• downlad.py: Files for downlading MNIST data sets
• ops.py: Operation functions
• utils.py: Functions dealing with images processing.

Prerequisites (my environments)

• Python 3.5.2
• Tensorflow > 0.14
• Numpy

Usage

First, download dataset with:

$ python download.py mnist

Second, write the main function with configuration you want.

Results

Result with same latent variables (0 to 100 epochs).

Each number (0 to 9) can be generated as you want by using CGAN.