Non-negative Matrix Factor Deconvolution; Extraction of Multiple Sound Sources from Monophonic Inputs. Smaragdis, P. In Puntonet, C. G. & Prieto, A., editors, Independent Component Analysis and Blind Signal Separation, of Lecture Notes in Computer Science, pages 494–499, Berlin, Heidelberg, 2004. Springer.
doi  abstract   bibtex   
In this paper we present an extension to the Non-Negative Matrix Factorization algorithm which is capable of identifying components with temporal structure. We demonstrate the use of this algorithm in the magnitude spectrum domain, where we employ it to perform extraction of multiple sound objects from a single channel auditory scene.
@inproceedings{smaragdis_non-negative_2004,
	address = {Berlin, Heidelberg},
	series = {Lecture {Notes} in {Computer} {Science}},
	title = {Non-negative {Matrix} {Factor} {Deconvolution}; {Extraction} of {Multiple} {Sound} {Sources} from {Monophonic} {Inputs}},
	isbn = {978-3-540-30110-3},
	doi = {10.1007/978-3-540-30110-3_63},
	abstract = {In this paper we present an extension to the Non-Negative Matrix Factorization algorithm which is capable of identifying components with temporal structure. We demonstrate the use of this algorithm in the magnitude spectrum domain, where we employ it to perform extraction of multiple sound objects from a single channel auditory scene.},
	language = {en},
	booktitle = {Independent {Component} {Analysis} and {Blind} {Signal} {Separation}},
	publisher = {Springer},
	author = {Smaragdis, Paris},
	editor = {Puntonet, Carlos G. and Prieto, Alberto},
	year = {2004},
	keywords = {Input Sound, Nonnegative Matrix, Nonnegative Matrix Factorization, Positive Matrix Factorization, Spectral Basis},
	pages = {494--499},
}

Downloads: 0