The language-independent bottleneck features
Authors:
- 1 December 2012
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 336-341
Abstract
In this paper we present novel language-independent bottleneck (BN) feature extraction framework. In our experiments we have used Multilingual Artificial Neural Network (ANN), where each language is modelled by separate output layer, while all the hidden layers jointly model the variability of all the source languages. The key idea is that the entire ANN is trained on all the languages simultaneously, thus the BN-features are not biased towards any of the languages. Exactly for this reason, the final BN-features are considered as language independent. In the experiments with GlobalPhone database, we show that Multilingual BN-features consistently outperform Monolingual BN-features. Also, cross-lingual generalization is evaluated, where we train on 5 source languages and test on 3 other languages. The results show that the ANN can produce very good BN-features even for unseen languages, in some cases even better than if we trained the ANN on the target language only.Keywords
This publication has 8 references indexed in Scilit:
- Region dependent linear transforms in multilingual speech recognition2012
- Convolutive Bottleneck Network features for LVCSR2011
- Study of probabilistic and Bottle-Neck features in multilingual environment2011
- Regularized subspace Gaussian mixture models for cross-lingual speech recognition2011
- On the use of a multilingual neural network front-end2008
- Probabilistic and Bottle-Neck Features for LVCSR of Meetings2007
- An evaluation of cross-language adaptation for rapid HMM development in a new language2002
- Tandem connectionist feature extraction for conventional HMM systems2002