Conference paper

 
Grézl, F., Karafiát, M.: Integrating recent MLP feature extraction techniques into TRAP architecture, In: Proceedings of Interspeech 2011, Florence, IT, ISCA, 2011, p. 1229-1232, ISBN 978-1-61839-270-1, ISSN 1990-9772
Publication language:english
Original title:Integrating recent MLP feature extraction techniques into TRAP architecture
Title (cs):Integrace moderních technik extrakce příznaků pomocí MLP do TRAP architektury
Pages:1229-1232
Proceedings:Proceedings of Interspeech 2011
Conference:Interspeech 2011
Place:Florence, IT
Year:2011
ISBN:978-1-61839-270-1
Journal:Proceedings of Interspeech, Vol. 2011, No. 8, FR
ISSN:1990-9772
Publisher:International Speech Communication Association
URL:http://www.fit.vutbr.cz/research/groups/speech/publi/2011/grezl_interspeech2011_204.pdf [PDF]
Keywords
TRAP processing, Bottle-Neck technique, subphoneme classes, LVCSR features
Annotation
The article shows the performance improvement and limitations of TRAP and HATS neural net systems for feature extraction in LVCSR when the bottel neck approach and phoneme states targets are introduced into them.
Abstract
This paper is focused on the incorporation of recent techniques for multi-layer perceptron (MLP) based feature extraction in Temporal Pattern (TRAP) and Hidden Activation TRAP (HATS) feature extraction scheme. The TRAP scheme has been origin of various MLP-based features some of which are now indivisible part of state-of-the-art LVCSR systems. The modifications which brought most improvement - sub-phoneme targets and Bottle-Neck technique - are introduced into original TRAP scheme. Introduction of sub-phoneme targets uncovered the hidden danger of having too many classes in TRAP/HATS scheme. On the other hand, Bottle-Neck technique improved the TRAP/HATS scheme so its competitive with other approaches.
BibTeX:
@INPROCEEDINGS{
   author = {František Grézl and Martin Karafiát},
   title = {Integrating recent MLP feature extraction techniques into
	TRAP architecture},
   pages = {1229--1232},
   booktitle = {Proceedings of Interspeech 2011},
   journal = {Proceedings of Interspeech},
   volume = {2011},
   number = {8},
   year = {2011},
   location = {Florence, IT},
   publisher = {International Speech Communication Association},
   ISBN = {978-1-61839-270-1},
   ISSN = {1990-9772},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php?id=9755}
}