Deep Learning for Virtual Tryon Clothes

Executive Summary

The goal of the project was to build a Virtual Tryon application.  Here the aim is to assist individuals to make the right purchase decisions by enabling them to virtually check out clothing on factors such as size, fit, and style. It involves mapping a virtual realistic representation of the clothing over an avatar to depict how the product would look. Also, there is the factor of ease in changing clothes and colors while exploring options.

The specific need of our client was an application that would enable virtually checking out of tops on an avatar i.e. while the avatar would remain fixed, the clothing could be changed at will. They wanted to sell this application to their customers who were mostly big ecommerce players involved in apparel merchandising.

About our Client

Client : Confidential

Location: USA

Industry: Publishing


Deep Learning, Image Segmentation, Image Processing, Pose Detection (OpenPose), PyTorch (Python) OpenCV, Pandas, NumPy, Matplotlib, Docker, CUDA

Virtual Tryon