Reverberation Modeling for Robust Speech Recognition

Maas R, Sehr A, Kellermann W (2011)


Publication Language: English

Publication Type: Conference contribution

Publication year: 2011

City/Town: Düsseldorf, Germany

Pages Range: 217-218

Conference Proceedings Title: Deutsche Jahrestagung für Akustik (DAGA)

Event location: Düsseldorf DE

Abstract

The REMOS (REverberation MOdeling for Speech recognition) concept for reverberation-robust distanttalking speech recognition [1] is presented in this paper. REMOS extends a conventional hidden Markov model (HMM) trained on close-talking data with a reverberation model describing the acoustical environment. The combination of both models is performed during recognition to match the reverberant observation. Since varying acoustic conditions only require a reestimation of the reverberation model, REMOS is significantly more flexible than recognition systems trained on reverberant data.

Authors with CRIS profile

How to cite

APA:

Maas, R., Sehr, A., & Kellermann, W. (2011). Reverberation Modeling for Robust Speech Recognition. In Deutsche Jahrestagung für Akustik (DAGA) (pp. 217-218). Düsseldorf, DE: Düsseldorf, Germany.

MLA:

Maas, Roland, Armin Sehr, and Walter Kellermann. "Reverberation Modeling for Robust Speech Recognition." Proceedings of the Deutsche Jahrestagung für Akustik (DAGA), Düsseldorf Düsseldorf, Germany, 2011. 217-218.

BibTeX: Download