This project retrieves thousands of photos of Eastern Bluebird. Then, a Generative Adversarial Network is trained to generate realistic bird photos that fool the discriminator.
This was made as an introduction to GANs and generative neural networks in general.
- PyTorch for the GAN model and training
- Flickr API for the pictures scraping
Open the main file of the repo, GAN - Bird Photos Generation, and click on "Open in Colab" at the top of the file. To avoid downloading the pictures, you'll have to import the dataset in your GDrive (send me a message for the links).