Real-Time Feedback Application for Musical Pitch Training in Deaf Individuals Using Knowledge of Results

Abstract

People who are deaf have difficulty singing due to pitch recognition issues, even when using hearing aids or cochlear implants. This study addresses this issue by developing an interactive training application that provides real-time visual feedback, including Knowledge of Results (KR)—external feedback indicating whether the sung pitch matches the intended note. When a user produces a sound, the application detects its pitch using RAPT (Robust Algorithm for Pitch Tracking) and displays the corresponding musical note. A real-time pitch slider allows users to monitor and adjust their pitch relative to the target note. This dual-feedback mechanism enables users to actively regulate their vocal output during practice. The system was tested in four singing training sessions with ten participants who are deaf. As a result, the participants were able to successfully sing “Twinkle Twinkle Little Star.” The training enabled them to improve pitch and volume control and to develop a foundational understanding of basic singing techniques. This application demonstrates the potential to provide effective singing training for people who are deaf who have traditionally been excluded from music education. By supporting self-regulated learning through visual and interactive feedback, the system offers a meaningful educational tool to promote inclusivity in music learning for the deaf community.