Chord Recognition in Symbolic Music: A Segmental CRF Model, Segment-Level Features, and Comparative Evaluations on Classical and Popular Music

Chord Recognition in Symbolic Music: A Segmental CRF Model, Segment-Level Features, and Comparative Evaluations on Classical and Popular Music

Blog Article

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 Cappuccino Mug 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 Western classical music and a newly created corpus of rock music.Experimental results show that the semi-CRF model performs substantially better than Intercooler previous approaches when trained on a sufficient number of labeled examples and remains competitive when the amount of training data is limited.