We use a video or a livestream from a webcam of a person dancing to control a fleet of Neatos (vacuuming robots). Our algorithm determines the keypoints (such a joints and facial features) on the person in the video and tracks those movements. Then, our Neatos, which represent various joints, recreate it!
Project Overview
Examine our system architecture to understand how each component seamlessly integrates. Explore the integration of joint detection with Neato movement and discover how a global map connects all these components.
Project Milestones
Learn about how we created the project and the design decisions made along the way! The project achieved three significant milestones: the implementation of joint detection, the orchestration of multiple robots, and the incorporation of odometry.
Potential Concerns
This robotic project, like any other, carries the potential for harm if misused. To ensure responsible usage, we have outlined guidelines for its operation and provided insights into how the project addresses potential negative societal impacts as outlined by CVPR (Conference on Computer Vision and Pattern Recognition).