Using Voice Segments to Improve Artist Classification of Music

Adam Berenzweig; Daniel P. W. Ellis; Steve Lawrence

Using Voice Segments to Improve Artist Classification of Music
Berenzweig, Adam
Ellis, Daniel P. W.
Lawrence, Steve
Electrical Engineering
Persistent URL:
Book/Journal Title:
Virtual synthetic and entertainment audio: proceedings of the AES 22nd international conference, 2002 June 15 - 17, Espoo, Finland
Book Author:
Huopaniemi, Jyri
Audio Engineering Society
Publisher Location:
New York
Is it easier to identify musicians by listening to their voices or their music? We show that for a small set of pop and rock songs, automatically-located singing segments form a more reliable basis for classification than using the entire track, suggesting that the singer‘s voice is more stable across different performances, compositions, and transformations due to audio engineering techniques than the instrumental background. The accuracy of a system trained to distinguish among a set of 21 artists improves by about 15% (relative to the baseline) when based on segments containing a strong vocal component, whereas the system suffers by about 35% (relative) when music-only segments are used. In another experiment on a smaller set, however, performance drops by about 35% (relative) when the training and test sets are selected from different albums, suggesting that the system is learning album-specific properties possibly related to audio production techniques, musical stylistic elements, or instrumentation, even when attention is directed toward the supposedly more stable vocal regions.
Electrical engineering
Artificial intelligence
Item views
text | xml
Suggested Citation:
Adam Berenzweig, Daniel P. W. Ellis, Steve Lawrence, , Using Voice Segments to Improve Artist Classification of Music, Columbia University Academic Commons, .

Columbia University Libraries | Policies | FAQ