Chord Recognition in Symbolic Music: A Segmental CRF Model, Segment-Level Features, and Comparative Evaluations on Classical and Popular Music
This work addresses harmonic analysis for music researchers and practitioners, offering an incremental improvement in chord recognition accuracy.
The authors tackled chord recognition in symbolic music by developing a semi-Markov Conditional Random Field model that jointly segments and labels chord spans, using segment-level features like purity and coverage. Results show it performs substantially better than previous methods with sufficient training data and remains competitive with limited data.
We present a new approach to harmonic analysis that is trained to segment music into a sequence of chord spans tagged with chord labels. Formulated as a semi-Markov Conditional Random Field (semi-CRF), this joint segmentation and labeling approach enables the use of a rich set of segment-level features, such as segment purity and chord coverage, that capture the extent to which the events in an entire segment of music are compatible with a candidate chord label. The new chord recognition model is evaluated extensively on three corpora of classical music and a newly created corpus of rock music. Experimental results show that the semi-CRF model performs substantially better than previous approaches when trained on a sufficient number of labeled examples and remains competitive when the amount of training data is limited.