/Conditional-Gans

The test code for Conditional convolutional adversarial networks.

Primary LanguagePythonMIT LicenseMIT

Conditional-Gans

The test code for Conditional Generative Adversarial Nets using tensorflow.

INTRODUCTION

Tensorflow implements of Conditional Generative Adversarial Nets.The paper should be the first one to introduce Conditional GANS.But they did not provide source codes.My code has some differences comparing the paper:The Gans is based on Convolution network and the code refer to DCGAN.

Prerequisites

  • tensorflow >=1.0

  • python 2.7

  • opencv 2.4.8

  • scipy 0.13

Usage

Download mnist:

$ python download.py mnist

Train:

$ python main.py --op 0

Test:

$ python main.py --op 1

Visualization:

$ python main.py --op 2

GIF:

$ python make_gif.py

Result on mnist

Visualization:

the visualization of weights:

the visualization of activation:

Reference code

DCGAN