Fun project for generating anime characters using Deep Convolutional Generative Adversarial Networks (DCGAN). Picture above shows some results generated by anime-GAN.
Implemented with:
- Python 2.7
- Tensorflow 1.13.1
- Keras 2.2.5
Dataset:
- Should be under a folder named: /data
- https://github.com/pavitrakumar78/Anime-Face-GAN-Keras/blob/master/anime_dataset_gen.py
- https://www.kaggle.com/aadilmalik94/animecharacterfaces
- Checkout https://github.com/nikitaa30/Manga-GAN
References:
Models:
- Discriminator and Generator:
Generates/Saves images per 200 epochs under /images and /models folders.