Abstract:

Regardless of how convenient shopping clothes online is, c ustomers frequently worry about how a piece of clothing present in the image might appear while making purchases of clothing .So, enabling customers to try on clothing virtually improve their buying experience, revolutionizing how people shop while also saving the money for merchants because of reduced product returns. Without utilizing any sort of 3D data, we described a Virtual Try-On which is image-based network. Through the use of a novel Geometric Matching Module, VTON first learns a thin-plate-spline transformation for tailoring the instore clothing to the target person's body shape. To achieve flawless integration of the warped clothing and produced image which increases the likelihood that the outcomes will be realistic, we used a Try-On Module that trains on a composition mask.

Description:

Virtual Try-On, Thin-Plate-spline, Composition mask, Convolutional Neural Network, Geometric Matching Module, Try-On Module