Paper Abstract and Keywords |
Presentation |
2022-05-13 10:45
Efficient DNN model for word lip-reading Daiki Arakane, Takeshi Saitoh (Kyutech) PRMU2022-4 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
This paper studies various deep learning models for lip-reading technology, including one of supervised learning of the video. Lip Reading in the Wild (LRW), one of the large-scale public datasets in lip-reading, is used for the recognition experiment. The recognition target of LRW is 500 English words, which was released in 2016. Initially, the recognition accuracy was 66.1%, but many research groups have been working on it, and the current SOTA has achieved 88.5% by 3D-Conv + ResNet18 + MS-TCN + knowledge distillation. This paper investigates effective deep learning models for lip-reading that combine WideResNet, EfficientNet, Transformer, Vision Transformer, regarding the SOTA model. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Lip-reading / word / deep neural network / LRW / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 122, no. 13, PRMU2022-4, pp. 18-23, May 2022. |
Paper # |
PRMU2022-4 |
Date of Issue |
2022-05-05 (PRMU) |
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 |
PRMU2022-4 |
Conference Information |
Committee |
PRMU IPSJ-CVIM |
Conference Date |
2022-05-12 - 2022-05-13 |
Place (in Japanese) |
(See Japanese page) |
Place (in English) |
Toyota Technological Institute |
Topics (in Japanese) |
(See Japanese page) |
Topics (in English) |
How to conduct research (post-graduation project for students) |
Paper Information |
Registration To |
PRMU |
Conference Code |
2022-05-PRMU-CVIM |
Language |
Japanese |
Title (in Japanese) |
(See Japanese page) |
Sub Title (in Japanese) |
(See Japanese page) |
Title (in English) |
Efficient DNN model for word lip-reading |
Sub Title (in English) |
|
Keyword(1) |
Lip-reading |
Keyword(2) |
word |
Keyword(3) |
deep neural network |
Keyword(4) |
LRW |
Keyword(5) |
|
Keyword(6) |
|
Keyword(7) |
|
Keyword(8) |
|
1st Author's Name |
Daiki Arakane |
1st Author's Affiliation |
Kyushu Institute of Technology, (Kyutech) |
2nd Author's Name |
Takeshi Saitoh |
2nd Author's Affiliation |
Kyushu Institute of Technology, (Kyutech) |
3rd Author's Name |
|
3rd Author's Affiliation |
() |
4th Author's Name |
|
4th Author's Affiliation |
() |
5th Author's Name |
|
5th Author's Affiliation |
() |
6th Author's Name |
|
6th Author's Affiliation |
() |
7th Author's Name |
|
7th Author's Affiliation |
() |
8th Author's Name |
|
8th Author's Affiliation |
() |
9th Author's Name |
|
9th Author's Affiliation |
() |
10th Author's Name |
|
10th Author's Affiliation |
() |
11th Author's Name |
|
11th Author's Affiliation |
() |
12th Author's Name |
|
12th Author's Affiliation |
() |
13th Author's Name |
|
13th Author's Affiliation |
() |
14th Author's Name |
|
14th Author's Affiliation |
() |
15th Author's Name |
|
15th Author's Affiliation |
() |
16th Author's Name |
|
16th Author's Affiliation |
() |
17th Author's Name |
|
17th Author's Affiliation |
() |
18th Author's Name |
|
18th Author's Affiliation |
() |
19th Author's Name |
|
19th Author's Affiliation |
() |
20th Author's Name |
|
20th Author's Affiliation |
() |
Speaker |
Author-1 |
Date Time |
2022-05-13 10:45:00 |
Presentation Time |
15 minutes |
Registration for |
PRMU |
Paper # |
PRMU2022-4 |
Volume (vol) |
vol.122 |
Number (no) |
no.13 |
Page |
pp.18-23 |
#Pages |
6 |
Date of Issue |
2022-05-05 (PRMU) |
|