Unsupervised Image To Image Translation

Unsupervised image-to-image translation is a task in the field of computer vision that aims to achieve image conversion without relying on paired real image data. This task learns the mapping relationship between images in different domains to realize cross-domain image generation, which has a wide range of application values, such as image style transfer and data augmentation.