FittingGAN: Fitting image Generation Based on Conditional Generative Adversarial Networks

Abstract

Recent studies have shown remarkable success in image generations using generative adversarial networks (GANs). However, how to deal with the fitting image generation, which is a task that generates a reasonable dressing image containing the input clothes is still an open problem. In this paper, we propose a condition generation model named FittingGAN which can achieve the generation of fitting scenes. The results show that It can generate fitting images with high resolution and realistic details, and FittingGAN have achieved good results in both qualitative and quantitative evaluations.