5-02 - Automatic Figured Bass Annotation Using the New Bach Chorales Figured Bass Dataset
Yaolong Ju, Sylvain Margot, Cory McKay, Luke Dahn, Ichiro Fujinaga
Keywords: Evaluation, datasets, and reproducibility, Novel datasets and use cases, Domain knowledge, Machine learning/Artificial intelligence for music, MIR fundamentals and methodology, Symbolic music processing, Musical features and properties, Harmony, chords, and tonality
Abstract:
This paper focuses on the computational study of figured bass, which remains an under-researched topic in MIR, likely due to a lack of machine-readable datasets. First, we introduce the Bach Chorales Figured Bass dataset (BCFB), a collection of 139 chorales composed by Johann Sebastian Bach that includes both the original music and figured bass annotations encoded in MusicXML, **kern, and MEI formats. We also present a comparative study on automatic figured bass annotation using both rule-based and machine learning approaches, which respectively achieved classification accuracies of 85.3% and 85.9% on BCFB. Finally, we discuss promising areas for MIR research involving figured bass, including automatic harmonic analysis.