Team 32: Machine Learning in Motion

The main purpose of this project is to capture the motion of individuals using a camera interface and connect their movements to interactive on-screen art elements. Our project utilizes real-time motion tracking using the built-in ZED2i camera object detection model and works on making the user interface smooth and reducing system lag as much as possible. In addition, it has artistic motion recognition for use in shows and for personal use. Renee Murray, our client, wants a version that is ready for production and easy to use. It has three-axis object tracking and a modern user interface.  

Presentation Video Link
Team Photo
Ambassador Profile
Team Poster
Problem Statement/Summary

As a dance instructor, Renee wanted to intertwine machine learning (ML) and art to create interactive art exhibits.​ She aimed to achieve this through motion capture; however, the first iteration used a less than ideal Motion Tracking Model.