Přístupnostní navigace
E-application
Search Search Close
Publication detail
BURGET, L., DUPONT, S., GARUDADRI, H., GRÉZL, F., HEŘMANSKÝ, H., JAIN, P., KAJAREKAR, S., MORGAN, N.
Original Title
QUALCOMM-ICSI-OGI Features for ASR
Type
conference paper
Language
English
Original Abstract
Our feature extraction module for the Aurora task is based on a combination of a conventional noise supression technique (Wiener filtering) with our temporal processing technigues (linear discriminant RASTA filtering and nonlinear TempoRAl Pattern (TRAP) classifier). We observe better than 58% relative error improvement on the prescribed Aurora Digit Task, a performance level that is somewhat better than the new ETSI Advanced Feature standard. Furthermore, to test generalization of our approach to an independent test set not available during development, we evaluate performance on American English SpeechDatCar digits and show 10.54% relative improvement over the new ETSI standard.
Keywords
feature extraction, distributed speech recognition, Aurora task, RASTA, TRAP
Authors
RIV year
2002
Released
20. 9. 2002
Publisher
International Speech Communication Association
Location
Denver
ISBN
1-876346-42-6
Book
Proc. 7th International Conference on Spoken Language Processing
Pages from
4
Pages to
7
Pages count
URL
http://www.fit.vutbr.cz/~burget/phd_activities/adami_icslp02.pdf
BibTex
@inproceedings{BUT10471, author="Lukáš {Burget} and Stephane {Dupont} and Harinath {Garudadri} and František {Grézl} and Hynek {Heřmanský} and Pratibha {Jain} and Sachin {Kajarekar} and Nelson {Morgan}", title="QUALCOMM-ICSI-OGI Features for ASR", booktitle="Proc. 7th International Conference on Spoken Language Processing", year="2002", pages="4", publisher="International Speech Communication Association", address="Denver", isbn="1-876346-42-6", url="http://www.fit.vutbr.cz/~burget/phd_activities/adami_icslp02.pdf" }