IEICE Technical Committee Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2022-03-27 15:40
Evaluation of Industrial Anomaly Detection using Diffusion Model
Yu Kashihara, Takashi Matsubara (Osaka Univ.) CCS2021-48
Abstract (in Japanese) (See Japanese page) 
(in English) Anomaly detection by generative models is achieved by comparing the reconstruction and the original image. However, existing generative models often lead to a blurred reconstruction and the loss of original image features (e.g., the orientation). In industrial anomaly detection, the objects often point to each direction and have detailed flaws. If the reconstruction is different from the original orientation and blurs the flaw, the anomaly detection fails. To avoid the problems, we propose the anomaly detection model based on diffusion model. The proposed model can reconstruct an image well and preserve original features. In this study, the model is evaluated on MVTeC AD, a dataset of industrial products anomaly detection, and demonstrates the area under receiver operating characteristic curve of 0.92. The score is significantly better than the existing generative models.
Keyword (in Japanese) (See Japanese page) 
(in English) deep learning / artificial intelligence / anomaly detection / generative model / diffusion model / / /  
Reference Info. IEICE Tech. Rep., vol. 121, no. 442, CCS2021-48, pp. 72-77, March 2022.
Paper # CCS2021-48 
Date of Issue 2022-03-20 (CCS) 
ISSN Online edition: ISSN 2432-6380
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 CCS2021-48

Conference Information
Committee CCS  
Conference Date 2022-03-27 - 2022-03-27 
Place (in Japanese) (See Japanese page) 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To CCS 
Conference Code 2022-03-CCS 
Language English (Japanese title is available) 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Evaluation of Industrial Anomaly Detection using Diffusion Model 
Sub Title (in English)  
Keyword(1) deep learning  
Keyword(2) artificial intelligence  
Keyword(3) anomaly detection  
Keyword(4) generative model  
Keyword(5) diffusion model  
1st Author's Name Yu Kashihara  
1st Author's Affiliation Osaka University (Osaka Univ.)
2nd Author's Name Takashi Matsubara  
2nd Author's Affiliation Osaka University (Osaka Univ.)
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 2022-03-27 15:40:00 
Presentation Time 25 minutes 
Registration for CCS 
Paper # CCS2021-48 
Volume (vol) vol.121 
Number (no) no.442 
Page pp.72-77 
Date of Issue 2022-03-20 (CCS) 

[Return to Top Page]

[Return to IEICE Web Page]

The Institute of Electronics, Information and Communication Engineers (IEICE), Japan