Publication Type:

Conference Paper

Source:

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, Dallas, TX, p.5010-5013 (2010)

ISBN:

9781424442966

URL:

http://www.scopus.com/inward/record.url?eid=2-s2.0-78049379903&partnerID=40&md5=66c81a1c8970bf233fbd6358b03180a7

Keywords:

Computational linguistics, Decoding, Hidden Markov models, Language identification, Language modeling, Multi-resolutions, Multilingual, Multilingual system, Mutual informations, Neural networks, Phone recognition, Signal processing, Speech recognition, Telephone sets

Abstract:

Phonotactic approach, phone recognition to be followed by language modeling, is one of the most popular approaches to language identification (LID). In this work, we explore how language identification accuracy of a phone decoder can be enhanced by varying acoustic resolution of the phone decoder, and subsequently how multiresolution versions of the same decoder can be integrated to improve the LID accuracy. We use mutual information to select the optimum set of phones for a specific acoustic resolution. Further, we propose strategies for building multilingual systems suitable for LID applications, and subsequently fine tune these systems to enhance the overall accuracy. ©2010 IEEE.

Notes:

cited By (since 1996)0; Conference of org.apache.xalan.xsltc.dom.DOMAdapter@5a0682a8 ; Conference Date: org.apache.xalan.xsltc.dom.DOMAdapter@616c11c4 Through org.apache.xalan.xsltc.dom.DOMAdapter@7e028cd3; Conference Code:81981

Cite this Research Publication

C. P. Sa Kumar, Li, Hb, Tong, Rbc, Matějka, Pd, Burget, Ld, and Černocký, Jd, “Tuning phone decoders for language identification”, in ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, Dallas, TX, 2010, pp. 5010-5013.