Paper Abstract and Keywords |
Presentation |
2023-03-02 10:55
Report on the 3rd Lip-Reading Challenge Takeshi Saitoh (Kyutech), Yuto Goto, Hiroyuki Nagano, Akihiro Kato, Masaki Nose (RICOH), Naoki Hiramoto (Mercoin), Tomohiro Hattori, Shiiya Aoyama, Yusuke Katoh, Ryuta Toshima, Takumi Nagawaki, Satoshi Tamura (Gifu University), Daiki Arakane (Kyutech) PRMU2022-76 IBISML2022-83 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
There is a machine lip-reading technology that uses a computer to estimate the utterance content using only visual information without using audio information. Despite more than 40 years of engineering analysis, it has not yet been put to practical use. Therefore, we planned the ``machine lip-reading challenge'' for the purpose of revitalizing the field of machine lip-reading research. When we held the 1st and 2nd machine lip-reading challenges in 2018 and 2019, we set the classification problem of 25 Japanese words as a task. Since high accuracy was achieved in the 2nd challenge, in 2022, the 3rd challenge, we held the problem of estimating the phoneme label sequence of Japanese sentences as a task. In this paper, we report on the overview of the competition, the database, the models of the baseline and the three participating teams, and the results. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Lip-reading / sentence / competition / / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 122, no. 404, PRMU2022-76, pp. 97-101, March 2023. |
Paper # |
PRMU2022-76 |
Date of Issue |
2023-02-23 (PRMU, IBISML) |
ISSN |
Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
PRMU2022-76 IBISML2022-83 |
|