This paper investigates the unsupervised adaptation of an acoustic model to a domain with mismatched acoustic conditions. We use techniques borrowed from the unsupervised training literature to adapt an acoustic model trained on the Wall Street Journal corpus to the Aurora-2 domain, which is composed of read digit strings over a simulated noisy telephone channel. We show that it is possible to use untranscribed in-domain data to get significant performance improvements, even when it is severely mismatched to the acoustic model.

PDF file

PostScript file

In: International Conference on Speech Communication and Technology

Publisher: International Speech Communication Association
© 2004 ISCA. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the ISCA and/or the author.

Details