Počet záznamů: 1
Improvements of Continuous Model for Memory-based Automatic Music Transcription
0347257 - UTIA-B 2011 RIV DK eng C - Konferenční příspěvek (zahraniční konf.)
Albrecht, Š. - Šmídl, Václav
Improvements of Continuous Model for Memory-based Automatic Music Transcription.
Proceedings of the 18th European signal processing conference. Aalborg: Eurasip, 2010, s. 487-491. ISSN 2076-1465.
[European signal processing conference. Aalborg (DK), 23.07.2010-27.07.2010]
Grant CEP: GA ČR GP102/08/P250
Výzkumný záměr: CEZ:AV0Z10750506
Klíčová slova: music transcription * extended Kalman filter
Kód oboru RIV: BD - Teorie informace
http://library.utia.cas.cz/separaty/2010/AS/smidl-improvements of continuous model for memory-based automatic music transcription.pdf
Automatic music transcription is a process recovering the most likely combination of sounds that produced the recorded audio signal. We are concerned with memory-based approach, where the observed signal is modeled as a superposition of sounds from a library. Moreover, we assume that only parts of the sounds can be played. The number of possible combinations is excessive and exact estimation is computationally prohibitive. We propose to transform the original discrete-event model into a less restricted parametrization and impose the constraints in a soft way via prior information. The resulting model is a non-linear state-space model with Gaussian disturbances. The posterior estimates are evaluated by the extended Kalman filter. Performance of the model is studied in simulation and it is shown that it outperforms previously published methods.
Trvalý link: http://hdl.handle.net/11104/0188070