Conference paper

GRÉZL František, KARAFIÁT Martin and VESELÝ Karel. Adaptation of Multilingual Stacked Bottle-neck Neural Network Structure for New Language. In: Proceedings of ICASSP 2014. Florencie: IEEE Signal Processing Society, 2014, pp. 7704-7708. ISBN 978-1-4799-2892-7.
Publication language:english
Original title:Adaptation of Multilingual Stacked Bottle-neck Neural Network Structure for New Language
Title (cs):Adaptace mutlilinguální vrstvené struktury neuronových sítí typu "bottle-neck" pro nový jazyk
Pages:7704-7708
Proceedings:Proceedings of ICASSP 2014
Conference:The 39th International Conference on Acoustics, Speech, and Signal Processing (ICASSP)
Place:Florencie, IT
Year:2014
ISBN:978-1-4799-2892-7
Publisher:IEEE Signal Processing Society
URL:http://www.fit.vutbr.cz/research/groups/speech/publi/2014/grezl_icassp2014_p7704_adapation.pdf [PDF]
Keywords
feature extraction, Bottle-neck features, neural network adaptation, multilingual neural networks, Stacked Bottle- Neck structure
Annotation
In this paper a multilingual training of Stacked Bottle- Neck neural network structure for feature extraction is addressed. While for languages with plentiful resources, the optimal approach is to train the BN-NN on the target data, limited resources call for re-using data from other languages.
Abstract
The neural network based features became an inseparable part of state-of-the-art LVCSR systems. In order to perform well, the network has to be trained on a large amount of in-domain data. With the increasing emphasis on fast development of ASR system on limited resources, there is an effort to alleviate the need of in-domain data. To evaluate the effectiveness of other resources, we have trained the Stacked Bottle-Neck neural networks structure on multilingual data investigating several training strategies while treating the target language as the unseen one. Further, the systems were adapted to the target language by re-training. Finally, we evaluated the effect of adaptation of individual NNs in the Stacked Bottle-Neck structure to find out the optimal adaptation strategy. We have shown that the adaptation can significantly improve system performance over both, the multilingual network and network trained only on target data. The experiments were performed on Babel Year 1 data.
BibTeX:
@INPROCEEDINGS{
   author = {Franti{\v{s}}ek Gr{\'{e}}zl and Martin Karafi{\'{a}}t and
	Karel Vesel{\'{y}}},
   title = {Adaptation of Multilingual Stacked Bottle-neck Neural
	Network Structure for New Language},
   pages = {7704--7708},
   booktitle = {Proceedings of ICASSP 2014},
   year = {2014},
   location = {Florencie, IT},
   publisher = {IEEE Signal Processing Society},
   ISBN = {978-1-4799-2892-7},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php?id=10556}
}

Your IPv4 address: 54.198.0.187
Switch to IPv6 connection

DNSSEC [dnssec]