Speech enhancement for robust speech recognition in car environments using griffiths-jim ANC based on two-paired microphones

Yongseok Cho, Hanseok Ko

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    7 Citations (Scopus)

    Abstract

    This paper presents a speech enhancement method which is to improve the speech recognition in car noisy environments. Conventional single and two channel algorithms differ from recognition performance according to each microphone's position (channel). Moreover, there is an ever-increasing demand for safety in cars, and speech recognition systems are nowadays often used. In this paper, we propose a method, hybrid algorithm using the Griffiths-Jim adaptive noise cancellation (ANC) based on two-paired microphones, which improved speech recognition result from reduced noise signal. The simulated experiments showed us that the proposed algorithm is effective in realized car environments, which is unrelated to channels in a car.

    Original languageEnglish
    Title of host publication2004 IEEE International Symposium on Consumer Electronics - Proceedings
    Pages123-127
    Number of pages5
    Publication statusPublished - 2004
    Event2004 IEEE International Symposium on Consumer Electronics - Proceedings - Reading, United Kingdom
    Duration: 2004 Sept 12004 Sept 3

    Publication series

    Name2004 IEEE International Symposium on Consumer Electronics - Proceedings

    Other

    Other2004 IEEE International Symposium on Consumer Electronics - Proceedings
    Country/TerritoryUnited Kingdom
    CityReading
    Period04/9/104/9/3

    Keywords

    • Adaptive noise cancellation (ANC)
    • Griffiths-Jim
    • Spectral subraction
    • Speech enhancements
    • Speech recognition

    ASJC Scopus subject areas

    • General Engineering

    Fingerprint

    Dive into the research topics of 'Speech enhancement for robust speech recognition in car environments using griffiths-jim ANC based on two-paired microphones'. Together they form a unique fingerprint.

    Cite this