Imago Justo Mapping using GAN

The Spanish word for “right” or “fair” is justo, while the Latin word for “image” is imago. Thus, the imago justo mapping should be used wherever possible when dealing with (textile) images (virtual try- on). There is a growing selection of fashions that are appropriate for both sexes. Online Halloween costume purchasing presents unique challenges for sorting and identifying various outfits. Picture segmentation using generative adversarial networks may be used to recreate a person's clothing if that were ever necessary. Two methods, generation and discrimination, allow for the reconstruction of the necessary part of the image. The GAN segmentation procedure employs a pair of networks. They may either be used as discriminators or as generators. To reduce the discriminator network's error rate, this generator is trained to create false data (images). But, if the consumer is interested in seeing how the clothing would appear on them, a virtual cloth try- on may be accomplished by mapping the fabric to the person. With these factors in mind, shoppers of the future may be able to design and order their perfect clothing image through the internet (size, color, fitting, etc.). It also has potential use in textile and clothing technology. In this research, we will employ a Generative Adversarial Neural Network to produce fabric pictures and build virtual dressing rooms in two dimensions (GAN). This is a Python-based project that leverages the PyTorch and gdown libraries and the Gradio UI deployment framework. Google Colab was the site of all of the project's development.