State-Of-The-Art of MIR System Using Gaussian Process

Abstract

:  MIR is a highly interdisciplinary field of  bridging the domains of digital audio signal processing, pattern recognition, software system design, and machine learning. MIR algorithms are capable of recognizing and extracting information, enabling systems to perform extensive sorting, searching, music recommendation, metadata generation, transcription, and even aiding/generating real-time performance.A lot of music data have become available recently either locally or over the Internet but in order for users to benefit from them, an efficient music information retrieval technology is necessary. Research in this area has been focused on tasks such as genre classification, artist identification, music mood estimation, cover song identification, music annotation, melody extraction, etc. which facilitate efficient music search and recommendation services, intelligent play-list generation and other attractive applications.

In this paper, we built two systems, one for music genre classification and another for music emotion recognition using both SVM and GP models, and finally compared their performances on two databases of similar size.