Muscle Memory is a collaborative project developed at Triple Innovation to explore the concept of learning new skills through repetitive motor movement guided by advanced body tracking and AI technology. The project initially focused on improving form in racket sports, such as perfecting serves or backhand returns, but its applications quickly expanded to include other disciplines like dance. By leveraging cutting-edge body tracking and motion analysis, Muscle Memory demonstrates how AI can accelerate skill acquisition in a variety of fields.
Racket Sports: Perfecting Form
Body Tracking with MetaQuest 3
This video demonstrates how the MetaQuest 3 headset tracks body movements from the waist up, accurately identifying stances like a backhand return or overhead lob shot.
Real-Time Feedback: Comparing to the Pros
Using Google MediaPipe for Motion Analysis
A demonstration of real-time feedback using Google MediaPipe. The video shows how a user's movements, captured via webcam, are compared to a professional's backhand motion, providing actionable feedback to improve form and timing.
Expanding to Dance: Teaching Complex Movements
Dance Training with Real-Time Feedback
This video showcases how users can upload a video of a dance move, such as a robot routine, and learn it through real-time motion tracking. The app compares the user’s movements with the professional’s and provides a QR code to download a personalized video with music.