Separating Underdetermined Convolutive Speech Mixtures

Michael Syskind Pedersen, DeLiang Wang, Jan Larsen, Ulrik Kjems

    Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

    183 Downloads (Pure)

    Abstract

    A limitation in many source separation tasks is that the number of source signals has to be known in advance. Further, in order to achieve good performance, the number of sources cannot exceed the number of sensors. In many real-world applications these limitations are too restrictive. We propose a method for underdetermined blind source separation of convolutive mixtures. The proposed framework is applicable for separation of instantaneous as well as convolutive speech mixtures. It is possible to iteratively extract each speech signal from the mixture by combining blind source separation techniques with binary time-frequency masking. In the proposed method, the number of source signals is not assumed to be known in advance and the number of sources is not limited to the number of microphones. Our approach needs only two microphones and the separated sounds are maintained as stereo signals.
    Original languageEnglish
    Title of host publicationICA 2006
    Volume3889
    PublisherSpringer Berlin / Heidelberg
    Publication date2006
    Pages674-681
    Publication statusPublished - 2006
    Event6th International Conference on Independent Component Analysis and Blind Signal Separation - Charleston, United States
    Duration: 5 Mar 20068 Mar 2006

    Conference

    Conference6th International Conference on Independent Component Analysis and Blind Signal Separation
    CountryUnited States
    CityCharleston
    Period05/03/200608/03/2006
    SeriesLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    ISSN0302-9743

    Cite this