Presentation 2004/5/21
Construction of Common Database "AURORA-2 J-AV/AURORA-3J-AV" for Evaluating Speech Recognition Method Under Noisy Environments
Daisuke NEGI, Toshiki MAENO, Takayuki KITASAKA, Kensaku MORI, Yasuhito SUENAGA, Chiyomi MIYAJIMA, Katsunobu ITOU, Kazuya TAKEDA, Fumitada ITAKURA, Yoshiki SANO, Yoshiki NINOMIYA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Researchers are having more attentions on automatic speech recognition under noisy environment using audio and video information together to improve recognition rates. Visual information may play a very important role in speech recognition since it is never affected by acoustic noises. However, it has not been fully used in existing actual speech recognition systems because there have been only a few large-scale bimodal databases. According to the specification of our common database named "AURORA-2J/AURORA-3J" for evaluating speech recognition method under noisy environments, we have built a new database "AURORA-2J-AV(indoor)/AURORA-3J-AV(in-vehicle)" by acquiring high quality color and near-infrared facial images in synchronization with aural signals. These databases contain "indoor" audiovisual data taken in a quiet room and "in-vehicle" audiovisual data acquired in a minivan while driving down the noisy streets. Since we plan to distribute the databases widely among researchers, we have been developing a new software framework to handle the databases quite easily.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Audiovisual automatic speech recognition / Multimedia database / AURORA
Paper # PRMU2004-24,MI2004-24,WIT2004-24
Date of Issue

Conference Information
Committee WIT
Conference Date 2004/5/21(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Well-being Information Technology(WIT)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Construction of Common Database "AURORA-2 J-AV/AURORA-3J-AV" for Evaluating Speech Recognition Method Under Noisy Environments
Sub Title (in English)
Keyword(1) Audiovisual automatic speech recognition
Keyword(2) Multimedia database
Keyword(3) AURORA
1st Author's Name Daisuke NEGI
1st Author's Affiliation Graduate School of Information Science, Nagoya University()
2nd Author's Name Toshiki MAENO
2nd Author's Affiliation Graduate School of Information Science, Nagoya University
3rd Author's Name Takayuki KITASAKA
3rd Author's Affiliation Graduate School of Information Science, Nagoya University
4th Author's Name Kensaku MORI
4th Author's Affiliation Graduate School of Information Science, Nagoya University
5th Author's Name Yasuhito SUENAGA
5th Author's Affiliation Graduate School of Information Science, Nagoya University
6th Author's Name Chiyomi MIYAJIMA
6th Author's Affiliation Graduate School of Information Science, Nagoya University
7th Author's Name Katsunobu ITOU
7th Author's Affiliation Graduate School of Information Science, Nagoya University
8th Author's Name Kazuya TAKEDA
8th Author's Affiliation Graduate School of Information Science, Nagoya University
9th Author's Name Fumitada ITAKURA
9th Author's Affiliation Graduate School of Information Science, Nagoya University
10th Author's Name Yoshiki SANO
10th Author's Affiliation Graduate School of Engineering, Nagoya University
11th Author's Name Yoshiki NINOMIYA
11th Author's Affiliation Faculty of Management Information Science, Nagoya University of Commerce & Business
Date 2004/5/21
Paper # PRMU2004-24,MI2004-24,WIT2004-24
Volume (vol) vol.104
Number (no) 93
Page pp.pp.-
#Pages 6
Date of Issue