DMCA
Experiments towards a Multi-language LVCSR Interface (2000)
| Venue: | in Int. Conf. on Spoken Language Processing |
| Citations: | 2 - 0 self |
BibTeX
@INPROCEEDINGS{Schultz00experimentstowards,
author = {Tanja Schultz and Alex Waibel},
title = {Experiments towards a Multi-language LVCSR Interface},
booktitle = {in Int. Conf. on Spoken Language Processing},
year = {2000},
pages = {129--132}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
This paper describes experiments towards a multilanguage human-computer speech interface. Our interface is designed for large vocabulary continuous speech input. For this purpose a multilingual dictation database has been collected under GlobalPhone, which is a project at the Interactive Systems Labs. This project investigates LVCSR systems in 15 languages of the world, namely Arabic, Chinese, Croatian, English, French, German, Italian, Japanese, Korean, Portuguese, Russian, Spanish, Swedish, Tamil, and Turkish. Based on a global phoneme set we build different multilingual speech recognizer and present several performance results in language independent and language adaptive setups. 1 Introduction With the distribution of speech technology products all over the world, multi-language humancomputer interfaces are of increasing importance. Beyond it, new methods for the fast adaptation to new target languages with only limited training data becomes a practical concern. Monolingual phone...
Keyphrases
multi-language lvcsr interface language adaptive setup limited training data monolingual phone multilingual dictation database multi-language humancomputer interface lvcsr system new method speech technology product multilanguage human-computer speech interface present several performance result global phoneme large vocabulary continuous speech input new target interactive system lab practical concern fast adaptation different multilingual speech recognizer
