Audio Genre Classification Using Rhythm and Bass-line Pattern Information

This paper discusses an approach for the feature extraction for audio genre classification and many other tasks of music information retrieval (MIR). Many musical genres are characterized not only by timbral information but also by temporal features such as rhythm patterns and bass-line patterns. In particular, modern music pieces mostly have certain fixed rhythm and bass-line patterns per genre, and the extraction of these representative patterns is thought to be useful for genre classification. We have proposed the extraction of rhythm and bass-line patterns, and using this method, a new feature extraction system was implemented into Marsyas which is a open source software framework for audio analysis, synthesis and retrieval for MIR.