Quantifying Musical Style: Ranking Symbolic Music based on Similarity to a Style
This work addresses the need for scalable and reproducible similarity evaluation in generative music systems and musicology, though it is incremental as it builds on existing computational methods.
The authors tackled the problem of computationally measuring musical similarity to a style, proposing StyleRank, which uses novel features and Random Forests to rank MIDI files based on similarity to a given style, achieving high correlation with human perception and enabling precise ranking of generated samples.
Modelling human perception of musical similarity is critical for the evaluation of generative music systems, musicological research, and many Music Information Retrieval tasks. Although human similarity judgments are the gold standard, computational analysis is often preferable, since results are often easier to reproduce, and computational methods are much more scalable. Moreover, computation based approaches can be calculated quickly and on demand, which is a prerequisite for use with an online system. We propose StyleRank, a method to measure the similarity between a MIDI file and an arbitrary musical style delineated by a collection of MIDI files. MIDI files are encoded using a novel set of features and an embedding is learned using Random Forests. Experimental evidence demonstrates that StyleRank is highly correlated with human perception of stylistic similarity, and that it is precise enough to rank generated samples based on their similarity to the style of a corpus. In addition, similarity can be measured with respect to a single feature, allowing specific discrepancies between generated samples and a particular musical style to be identified.