The Analysis and Comparison of Vital Acoustic Features in Content-Based Classification of Music Genre

Digital music is becoming increasingly popular in the Internet, and content-based musical genre classification has gained significant attentions in the field of musical retrieval. In this paper, the acoustic musical features are extracted from the viewpoints of both signal processing and the musical dimension. By comparing the performance of classifier of different combination of acoustic features, the contributions of corresponding features are evaluated. Finally, timbre and tonality feature sets are found to be the most effective features in music genre recognition.