Be-Yoncé is an augmented reality webcam application that allows users to virtually try on Beyoncé's iconic clothing while playing music and taking downloadable screenshots. It was built using Tensorflow.js' PoseNet, a machine learning model which estimates human movement and pose in real time. We adapted PoseNet for a React/Redux frontend, and we built out the backend using Node, Express, PostgreSQL, and Sequelize. The clothing items were dynamically positioned on an HTML5 canvas using utility functions that allow them to move and scale with the user’s motion.
Team Members: Chelsi Asulin, Alex Grazier, Macarena Carreno, Linda Saraguro