This repo contains the Colab notebooks for the GAN series on PyImageSearch which is many people's go-to place for learning about computer vision, deep learning and OpenCV. These posts are also part of the PyImageSearch University courses where you can access both the tutorials and videos. The GAN implementations are written in TensorFLow 2 and Keras.
| Blog Post Title | Tweet | Tutorial | Open in Colab |
|---|---|---|---|
| 1. Intro GANs | Link | Link | N/A |
| 2. Get Started: DCGAN for Fashion-MNIST | Link | Link | Link |
| 3. GAN Training Challenges: DCGAN for Color Images | Link | Link | Link |
| 4. Anime Faces with WGAN and WGAN-GP | Link | Link | WGAN & WGAN-GP |
Read my Intro to GANs tutorial to learn how GANs work and a brief intro to the GAN variants and applications.
Read Get Started: DCGAN for Fashion-MNIST to learn how to implement a DCGAN to generate gray-scale (28x28x1) Fashion-MNIST like images.
Read this GAN Training Challenges: DCGAN for Color Images to learn how to use DCGAN to generate color (64x64x3) fashion images.
In Anime Faces with WGAN and WGAN-GP, I discuss how to move from DCGAN to WGAN, and then to WGAN-GP, for generating anime faces.
- WGAN notebook
- WGAN-GP notebook
Please note that any content, artwork or code in this repo is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, which means you are free to share and adapt it, under the condition that you give appropriate credit to the author and you may not use it for commercial purposes.