Fast and flexible Kullback-Leibler divergence based acoustic modeling for non-native speech recognition
| Type of publication: | Idiap-RR |
| Citation: | Imseng_Idiap-RR-01-2012 |
| Number: | Idiap-RR-01-2012 |
| Year: | 2012 |
| Month: | 1 |
| Institution: | Idiap |
| Abstract: | One of the main challenge in non-native speech recognition is how to handle acoustic variability present in multiaccented non-native speech with limited amount of training data. In this paper, we investigate an approach that addresses this challenge by using Kullback-Leibler divergence based hidden Markov models (KL-HMM). More precisely, the acoustic variability in the multi-accented speech is handled by using multilingual phoneme posterior probabilities, estimated by a multilayer perceptron trained on auxiliary data, as input feature for the KL-HMM system. With limited training data, we then build better acoustic models by exploiting the advantage that the KL-HMM system has fewer number of parameters. On HIWIRE corpus, the proposed approach yields a performance of 1.9% word error rate (WER) with 149 minutes of training data and a performance of 5.5% WER with 2 minutes of training data. |
| Keywords: | Hidden Markov Model, Kullback-Leibler divergence, multilayer perceptron, Posterior features |
| Projects |
Idiap IM2 |
| Authors | |
| Crossref by |
Imseng_ASRU_2011 |
| Added by: | [ADM] |
| Total mark: | 0 |
|
Attachments
|
|
|
|
|
Notes
|
|
|
|
|
