Can computers learn what makes the most iconic jazz musicians stand out?
Machine learning models can identify jazz pianists from recordings and reveal the musical "fingerprints" that make individual performers recognizable, according to a study investigating 20 famous jazz pianists. The research, published in Nature Machine Intelligence, could offer insight into artist attribution and style, cultural heritage, and music education.
Researchers have developed machine learning models capable of identifying jazz pianists from recordings with high accuracy, revealing the unique musical fingerprints that distinguish individual performers. The study, published in Nature Machine Intelligence, trained models on 84 hours of recordings from 1,629 performances by 20 famous jazz pianists, converting the audio into MIDI piano roll format for analysis.
The best-performing model achieved 94.4% accuracy in identifying performers, with harmonic features proving most informative, followed by rhythm and melody. While the research offers insights into artist attribution and style, the authors note that MIDI representations cannot capture all aspects of jazz performance, such as timbral qualities. Future work could extend the approach to other genres, instruments, and underrepresented musicians.
Brief written by urgent.news from Phys.org's own syndicated text. Machine-written — may contain errors; check the original before relying on it.
