%0 Thesis %1 olsson2007algorithms %A Olsson, Rasmus Kongsgaard %D 2007 %K audio separation source %T Algorithms for Source Separation - with Cocktail Party Applications %X n this thesis, a number of possible solutions to source separation are suggested. Although they differ significantly in shape and intent, they share a heavy reliance on prior domain knowledge. Most of the developed algorithms are intended for speech applications, and hence, structural features of speech have been incorpo- rated. Single-channel separation of speech is a particularly challenging signal process- ing task, where the purpose is to extract a number of speech signals from a single observed mixture. I present a few methods to obtain separation, which rely on the sparsity and structure of speech in a time-frequency representation. My own contributions are based on learning dictionaries for each speaker separately and subsequently applying a concatenation of these dictionaries to separate a mixture. Sparse decompositions required for the decomposition are computed using non- negative matrix factorization as well as basis pursuit. In my work on the multi-channel problem, I have focused on convolutive mix- tures, which is the appropriate model in acoustic setups. We have been successful in incorporating a harmonic speech model into a greater probabilistic formula- tion. Furthermore, we have presented several learning schemes for the parameters of such models, more specifically, the expectation-maximization (EM) algorithm and stochastic and Newton-type gradient optimization.

@mastersthesis{olsson2007algorithms, abstract = {n this thesis, a number of possible solutions to source separation are suggested. Although they differ significantly in shape and intent, they share a heavy reliance on prior domain knowledge. Most of the developed algorithms are intended for speech applications, and hence, structural features of speech have been incorpo- rated. Single-channel separation of speech is a particularly challenging signal process- ing task, where the purpose is to extract a number of speech signals from a single observed mixture. I present a few methods to obtain separation, which rely on the sparsity and structure of speech in a time-frequency representation. My own contributions are based on learning dictionaries for each speaker separately and subsequently applying a concatenation of these dictionaries to separate a mixture. Sparse decompositions required for the decomposition are computed using non- negative matrix factorization as well as basis pursuit. In my work on the multi-channel problem, I have focused on convolutive mix- tures, which is the appropriate model in acoustic setups. We have been successful in incorporating a harmonic speech model into a greater probabilistic formula- tion. Furthermore, we have presented several learning schemes for the parameters of such models, more specifically, the expectation-maximization (EM) algorithm and stochastic and Newton-type gradient optimization.}, added-at = {2011-09-30T14:06:50.000+0200}, author = {Olsson, Rasmus Kongsgaard}, biburl = {https://www.bibsonomy.org/bibtex/264c40230ce9a0dfed59ffd5c2bcddb58/nosebrain}, interhash = {551fc83a10baf6a81e555f01ca5297b3}, intrahash = {64c40230ce9a0dfed59ffd5c2bcddb58}, keywords = {audio separation source}, timestamp = {2012-11-04T17:54:17.000+0100}, title = {Algorithms for Source Separation - with Cocktail Party Applications}, year = 2007 }