­
­
­
­

Inventi Impact: Audio, Speech & Music Processing

Articles

  • Inventi:easm/41/13
    A PERCEPTUAL MASKING APPROACH FOR NOISE ROBUST SPEECH RECOGNITION
    30-Jun-2013 Research 2013 : July - September
    Hari Krishna Maganti, Marco Matassoni

    This article describes a modified technique for enhancing noisy speech to improve automatic speech recognition\r\n(ASR) performance. The proposed approach improves the widely used spectral subtraction which inherently suffers\r\nfrom the associated musical noise effects. Through a psychoacoustic masking and critical band variance normalization\r\ntechnique, the artifacts produced by spectral subtraction are minimized for improving the ASR accuracy. The popular\r\nadvanced ETSI-2 front end is tested for comparison purposes. The performed speech recognition evaluations on the\r\nnoisy standard AURORA-2 tasks show enhanced performance for all noise conditions.

    How to Cite this Article
    Cite this article as: Maganti and Matassoni: A perceptual masking approach\r\nfor noise robust speech recognition. EURASIP Journal on Audio, Speech, and\r\nMusic Processing 2012 2012:29.\r\nOpen Access\r\nunder Creative Commons Attribution Licence\r\nCC Compliant Citation: Maganti and Matassoni: A perceptual\r\nmasking approach for noise robust speech recognition.\r\nEURASIP Journal on Audio, Speech, and Music Processing\r\n2012 2012:29. doi:10.1186/1687-4722-2012-29.
    Download Full Text