This project uses a conditional Generative Adversarial Network (cGAN) to colorize anime-style sketches. The model architecture is based on the Pix2Pix paper, which is effective for image-to-image translation tasks.
Latest commits.
Builders behind this project.