Gain-robust multi-pitch tracking using sparse nonnegative matrix factorization

Robert Peharz, Michael Wohlmayr, Franz Pernkopf

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdrageAcademicpeer review

6 Citaten (Scopus)

Samenvatting

While nonnegative matrix factorization (NMF) has successfully been applied for gain-robust multi-pitch detection, a method to track pitch values over time was not provided. We embed NMF-based pitch detection into a recently proposed pitch-tracking system, based on a factorial hidden Markov model (FHMM). The original system models speech spectra with Gaussian mixture models, which is sensitive to a gain mismatch between training and test data. We therefore combine the advantages of these two approaches and derive a gain-adaptive observation model for the FHMM. As training algorithm we use a modification of ℓ0-sparse NMF, which represents the short-time spectrum with scalable basis vectors. In experiments we show that the new approach significantly increases the gain-robustness of the original tracking system.

Originele taal-2Engels
Titel2011 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2011 - Proceedings
UitgeverijInstitute of Electrical and Electronics Engineers
Pagina's5416-5419
Aantal pagina's4
ISBN van geprinte versie9781457705397
DOI's
StatusGepubliceerd - 18 aug. 2011
Extern gepubliceerdJa
Evenement2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2011) - Prague, Tsjechië
Duur: 22 mei 201127 mei 2011
Congresnummer: 36

Congres

Congres2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2011)
Verkorte titelICASSP 2011
Land/RegioTsjechië
StadPrague
Periode22/05/1127/05/11
Ander36th International Conference on Acoustics, Speech and Signal Processing

Vingerafdruk

Duik in de onderzoeksthema's van 'Gain-robust multi-pitch tracking using sparse nonnegative matrix factorization'. Samen vormen ze een unieke vingerafdruk.

Citeer dit