We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Instantaneous model adaptation method for reverberant speech recognition.
- Authors
Sung Min Ban; Hyung Soon Kim
- Abstract
An acoustic model adaptation algorithm is proposed for reverberant speech recognition. Inspired by the eigenvoice adaptation framework, multiple acoustic models reflecting various reverberant environments are combined for instantaneous adaptation. Using artificially generated reverberant speech, multiple acoustic models are trained according to multiple reverberation times. The mean vectors of the optimal acoustic model are obtained as a weighted sum of those of multiple acoustic models by using a maximum-likelihood criterion. For effective model combination, reverberant speech is preprocessed. Experiments on English continuous speech recognition tasks in a simulated reverberant environment show that the proposed method performs better than the conventional adaptation techniques.
- Subjects
ACOUSTIC models; SPEECH perception; ENGLISH language; SPEECH processing systems; NORMAL voice
- Publication
Electronics Letters (Wiley-Blackwell), 2015, Vol 51, Issue 6, p528
- ISSN
0013-5194
- Publication type
Article
- DOI
10.1049/el.2014.4152