A Discriminative Model for Polyphonic Piano Transcription. (29th October 2006)

• Record Type: Journal Article
• Authors: Poliner Poliner, Graham E. Graham E.; Ellis Ellis, Daniel P. W. Daniel P. W.
• Other Names: Goto Goto Masataka Masataka Academic Editor.
• Is Part Of: EURASIP journal on advances in signal processing. Volume 2007(2007)
• Publication Date: 2006-10-29
• Subjects: Signal processing -- Periodicals; Traitement du signal; Signal processing; Periodicals; 621.3822
• DOI: 10.1155/2007/48317
• Languages: English
• ISSNs: 1687-6172

Abstract

Abstract : We present a discriminative model for polyphonic piano transcription. Support vector machines trained on spectral features are used to classify frame-level note instances. The classifier outputs are temporally constrained via hidden Markov models, and the proposed system is used to transcribe both synthesized and real piano recordings. A frame-level transcription accuracy of 68% was achieved on a newly generated test set, and direct comparisons to previous approaches are provided.