AnimeFaceGen is a project that utilizes Generative Adversarial Networks (GANs) to generate high-quality anime-style faces. GANs are a class of artificial intelligence algorithms used in unsupervised machine learning, implemented via a system of two neural networks competing against each other in a zero-sum game framework.
This repository contains code for training and generating anime faces using GANs. The generated faces exhibit diverse characteristics including varying facial features, expressions, and styles, providing a rich dataset for various creative and research purposes.
Make sure you have the following dependencies installed to run the project:
- numpy==1.19.5
- pandas==1.3.3
- tensorflow==2.7.0
- keras==2.7.0
- matplotlib==3.4.3
- opencv-python==4.5.3.56
You can install them using pip install -r requirements.txt.
Generative Adversarial Networks (GANs) consist of two neural networks: a generator and a discriminator.
- Generator: The generator network takes random noise as input and generates images. Its goal is to generate images that are indistinguishable from real images.
- Discriminator: The discriminator network takes both real images from the dataset and fake images generated by the generator as input. Its goal is to distinguish between real and fake images.
During training, the generator tries to generate realistic images to fool the discriminator, while the discriminator learns to better distinguish between real and fake images. This adversarial process leads to the generator producing increasingly realistic images over time.
-
Training: Use the provided scripts to train the GAN model on a dataset of anime faces. Adjust hyperparameters and model architecture as needed.
-
Generation: After training, use the trained model to generate anime faces by providing random noise as input to the generator network.
Above are some examples of anime faces generated using the trained GAN model.