Artificial Intelligence-based Musical Instrument Accompaniment System
Kevin P. Chen · Applied and Computational Engineering · 2023
The string learning process requires students to be able to accurately grasp the accuracy of notes and rhythms in the process of playing, and its evaluation is usually done by instructors, but long one-on-one instruction is difficult to achieve in actual teaching. In this project, through the research of audio hardware systems and digital signal processing software technology, I design a string performance recognition robot system combining hardware and software, to realize automatic accompanying practice. The hardware of this system consists of a Raspberry Pi card-type computer, recording equipment, and cueing equipment. The software system processes the audio signal collected by the microphone through artificial intelligence technology and digital signal processing technology to realize the recognition of notes, intensity, and other elements according to the frequency characteristics of string instruments. The accuracy of the user's performance is judged by comparing the recognition result with the content of the score, and the user is guided by real-time hints through the display system. The experimental results show that the system can recognize 99% of the string audio with good recognition stability, so it can give accurate and timely performance guidance to the player.