This repository is the final result of the project on Machine Learning and the Applications at Yonsei University.
Source: Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
The previous image-to-image translation such as Pix2Pix learns the mapping between an input image and an output image with a set of aligned image pairs. However, for many tasks, paired training dataset is unavailable. CycleGAN solves this issue by learning to translate an image from a source domain X to a target domain Y in the absence of paired examples.
- Consists of 20,000+ face images (Only single face in one image)
- Correspondingly aligned & cropped faces
- Images are labelled by age and gender
Detailed description with PDF file @ Face Aging using CycleGAN.pdf