Riedhammer KT, Bocklet T, Orozco-Arroyave JR, Nöth E (2014)
Publication Status: Published
Publication Type: Conference contribution, Conference Contribution
Publication year: 2014
Publisher: Institute of Electrical and Electronics Engineers Inc.
Article Number: 6926034
ISBN: 9783800736409
URI: https://www.scopus.com/inward/record.url?partnerID=HzOxMe3b&scp=84939530520&origin=inward
In this article, we describe a semi-automatic calibration algorithm for dereverberation by spectral subtraction. We verify the method by a comparison to a manual calibration derived from measured room impulse responses (RIR). We conduct extensive experiments to understand the effect of all involved parameters and to verify values suggested in the literature. The experiments are performed on a text read by 31 speakers and recorded by a headset and three far-field microphones. Results are measured in terms of automatic speech recognition (ASR) performance using a 1-gram model to emphasize acoustic recognition performance. To accommodate for the acoustic change by dereverberation we apply supervised MAP adaptation to the hidden Markov model output probabilities. The combination of dereverberation and adaptation yields a relative improvement of about 35% in terms of word error rate (WER) compared to the original signal.
APA:
Riedhammer, K.T., Bocklet, T., Orozco-Arroyave, J.R., & Nöth, E. (2014). Semi-automatic calibration for dereverberation by spectral subtraction for continuous speech recognition. In Proceedings of the 11th ITG Symposium on Speech Communication. Institute of Electrical and Electronics Engineers Inc..
MLA:
Riedhammer, Korbinian Thomas, et al. "Semi-automatic calibration for dereverberation by spectral subtraction for continuous speech recognition." Proceedings of the 11th ITG Symposium on Speech Communication Institute of Electrical and Electronics Engineers Inc., 2014.
BibTeX: Download