Chenxi Huang¹ and Rodrigo Onate², ¹USA, ²California State Polytechnic University, USA

This project aims to make dance learning more accessible for deaf individuals through a program called RhythmSense. Many people who are deaf struggle to follow rhythm or music during dance. My solution combines AI-based pose detection, sound wave visualization, and a vibration feedback band. The system uses MediaPipe to track movements, compares them to a reference video, and provides instant visual and physical feedback. During testing, I focused on improving accuracy, reducing vibration delay, and ensuring that the program worked in different lighting and motion conditions. The results showed that the app could identify mistakes and match rhythm effectively. Overall, RhythmSense helps dancers not only see their errors but also feel the beat through vibration. This technology creates a more inclusive way for everyone, including deaf users, to experience and enjoy dance.

Dance learning, Pose estimation, Real-time feedback, Vibration Band