Title
Degenerate Unmixing Estimation Technique using the Constant Q Transform
Abstract
The Degenerate Unmixing Estimation Technique (DUET) is a Blind Source Separation (BSS) algorithm for stereo audio. DUET depends on an amplitude-phase 2d histogram built from the differences between the two channels, where peaks in the histogram indicate sources in the mixture. If peaks overlap, separation becomes unfeasible. This is often the case for music mixtures. We propose to improve peak separation by building histograms from time-frequency representations based on the Constant Q Transform (CQT) instead of the Fourier Transform (FT). The CQT has a logarithmic frequency resolution matching the geometrically spaced notes of the Western music scale. We also adaptively resize histogram bins and use Wiener filtering to improve peak resolving and source reconstruction. Results on mixtures of harmonic musical instruments show improvement in separation, especially at low frequencies and for closely spaced sources.
Year
DOI
Venue
2011
10.1109/ICASSP.2011.5946379
Acoustics, Speech and Signal Processing
Keywords
Field
DocType
Wiener filters,audio signal processing,blind source separation,channel estimation,harmonic analysis,music,time-frequency analysis,DUET,Fourier transform,Western music scale,Wiener filtering,adaptively resize histogram bins,amplitude-phase 2D histogram,blind source separation algorithm,constant Q transform,geometrically spaced note,harmonic musical instrument,histogram building,logarithmic frequency resolution matching,music mixture,peak separation,source reconstruction,spaced sources,stereo audio,time-frequency representation,unmixing estimation technique,Blind Source Separation,Constant Q Transform,Degenerate Unmixing Estimation Technique
Wiener filter,Histogram,Constant Q transform,Pattern recognition,Fourier transform,Harmonic analysis,Time–frequency analysis,Artificial intelligence,Blind signal separation,Source separation,Mathematics
Conference
ISSN
ISBN
Citations 
1520-6149 E-ISBN : 978-1-4577-0537-3
978-1-4577-0537-3
1
PageRank 
References 
Authors
0.38
4
2
Name
Order
Citations
PageRank
Zafar Rafii116310.92
Bryan Pardo283063.92