A Review of the Research on the Combination of Artificial Intelligence and Instrument Recognition

Abstract

Artificial intelligence has great potential value in the music industry. From the perspective of music recognition based on machine learning, there are numerous possibilities in the industry of music and computer science intersection, music composition and generation; music analysis and preprocessing; music education and interaction, etc. After elaborating on music analysis and preprocessing, a potent about musical instrument recognition can be done with artificial intelligence have discovered. By clarifying previous research combining traditional and modern music features extraction methods like mel-frequency cepstral coefficients (MFCC) and attention mechanism, proposed based on music theory, with deep learning models like convolutional neural network (CNN), artificial neural network (ANN), and Keras for musical instrument recognition. By giving suggestions in this article, musical instrument recognition models’ efficacy can become more precise. Therefore, in the future, researchers can have the possibility to dig into deeper and more branches of the intersection industry, resulting in more unexpected value.