ITE Technical Group Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2022-03-08 13:28
Style-invariant representations and frame-aware constraints for manga face image clustering
Shunta Komatsu (TUS), Ryosuke Furuta (UTokyo), Yukinobu Taniguchi (TUS), Ryota Hinami, Shonosuke Ishiwatari (Mantra)
Abstract (in Japanese) (See Japanese page) 
(in English) Character recognition in manga is an important task to improve the usability of e-comics through recommendation systems and automatic translation systems.
However, character identification in manga is difficult because the style of manga varies greatly from author to author and work to work.
In this paper, we propose a clustering method for manga face images, which is robust for unknown authors and works.
In particular, we show that training the model using domain generalization with self-supervised learning enables effective feature extraction even for unknown styles.
We also show that the constraints based on domain-specific knowledge in manga improve clustering accuracy.
As a result of experiments on a manga image dataset, we confirmed that the proposed method improves the clustering accuracy.
Keyword (in Japanese) (See Japanese page) 
(in English) Clustering / Deep Learning / Manga / / / / /  
Reference Info. ITE Tech. Rep., vol. 46, pp. 153-156, March 2022.
Paper #  
Date of Issue 2022-03-01 (AIT) 
ISSN Print edition: ISSN 1342-6893    Online edition: ISSN 2424-1970
Download PDF

Conference Information
Conference Date 2022-03-08 - 2022-03-08 
Place (in Japanese) (See Japanese page) 
Place (in English) Online 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Expressive Japan 2022 
Paper Information
Registration To IIEEJ 
Conference Code 2022-03-AIT-IIEEJ-AS-ARTS 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Style-invariant representations and frame-aware constraints for manga face image clustering 
Sub Title (in English)  
Keyword(1) Clustering  
Keyword(2) Deep Learning  
Keyword(3) Manga  
1st Author's Name Shunta Komatsu  
1st Author's Affiliation Tokyo University of Science (TUS)
2nd Author's Name Ryosuke Furuta  
2nd Author's Affiliation The University of Tokyo (UTokyo)
3rd Author's Name Yukinobu Taniguchi  
3rd Author's Affiliation Tokyo University of Science (TUS)
4th Author's Name Ryota Hinami  
4th Author's Affiliation Mantra Inc. (Mantra)
5th Author's Name Shonosuke Ishiwatari  
5th Author's Affiliation Mantra Inc. (Mantra)
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 2022-03-08 13:28:00 
Presentation Time 12 minutes 
Registration for IIEEJ 
Paper # AIT2022-78 
Volume (vol) vol.46 
Number (no) no.10 
Page pp.153-156 
Date of Issue 2022-03-01 (AIT) 

[Return to Top Page]

[Return to ITE Web Page]

The Institute of Image Information and Television Engineers (ITE), Japan