Music tempo classification using audio spectrum centroid, audio spectrum flatness, and audio spectrum spread based on MPEG-7 audio features

Abstract

Music has become an integral part in human life. Recent studies show that music can affect human's mood. For example, music with slow tempo will cause the listener feel relaxed. Meanwhile, music with fast tempo will cause the listener feel excited. This paper discusses about music tempo classification using features from MPEG-7 based on Support Vector Machine (SVM). MPEG-7 is international standardized multimedia metadata in ISO/IEC 15938. The audio features used in this experiment are Audio Spectrum Centroid, Audio Spectrum Flatness, and Audio Spectrum Spread. Features are classified using SVM. The result of this operation is a classification of music based on its beats-per-minute (BPM). The classification rate of the experiment is 80%.