Information: Join today and make your research activities more affordable! Technical workshop participation fees and annual registration fees are available at member rates.
Notice: [Important] Announcement of Changes to Registration Fee Payment and Manuscript Upload Procedures for IEICE Technical Meetings
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 2026-07-03 11:25
Noise-Aware Deep Learning for Drosophila 3D Image Classification
Md. Humaun Kabir (Grad. School of CSE, Hokkaido University), Md. Al Mehedi Hasan (RIES, Hokkaido Univ.), Walker Peterson (Dept. of Chemistry, The Univ. of Tokyo), Koji Tabata (RIES, Hokkaido Univ.), Masahiro Sonoshita (Institute for Genetic Medicine, Hokkaido Univ.), Keisuke Goda (Dept. of Chemistry, The Univ. of Tokyo), Tamiki Komatsuzaki (RIES, Hokkaido Univ.) NC2026-10 IBISML2026-10
Abstract (in Japanese) (See Japanese page) 
(in English) Deep neural networks often assume clean labels and confident predictions. However, real-world datasets frequently contain label noise, ambiguity, and low-quality samples, leading to overconfident errors. This issue is particularly critical in safety-sensitive applications such as medical diagnosis, drug screening, autonomous driving, and financial forecasting, where misclassification can have serious consequences. This study addresses robust learning from noisy and ambiguous data and reliable decision-making under uncertainty. A pretrained ResNet-18 is integrated with several noise- and uncertainty-aware selective prediction methods, including SelectiveNet, RISAN, and Learning-to-Defer. Experimental results show that RISAN outperforms the other methods, improving robustness and decision reliability in safety-critical and high-stakes applications.
Keyword (in Japanese) (See Japanese page) 
(in English) Reliable AI / Abstention / Learning-to-Abstain / Learning-to-Defer / Noisy Labels / Selective Prediction / /  
Reference Info. IEICE Tech. Rep., vol. 126, no. 91, IBISML2026-10, pp. 38-45, July 2026.
Paper # IBISML2026-10 
Date of Issue 2026-06-25 (NC, 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 NC2026-10 IBISML2026-10

Conference Information
Committee NC IBISML IPSJ-BIO IPSJ-MPS  
Conference Date 2026-07-01 - 2026-07-03 
Place (in Japanese) (See Japanese page) 
Place (in English) OIST 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To IBISML 
Conference Code 2026-07-NC-IBISML-BIO-MPS 
Language English 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Noise-Aware Deep Learning for Drosophila 3D Image Classification 
Sub Title (in English)  
Keyword(1) Reliable AI  
Keyword(2) Abstention  
Keyword(3) Learning-to-Abstain  
Keyword(4) Learning-to-Defer  
Keyword(5) Noisy Labels  
Keyword(6) Selective Prediction  
Keyword(7)  
Keyword(8)  
1st Author's Name Md. Humaun Kabir  
1st Author's Affiliation Graduate School of Chemical Sciences and Eng., Hokkaido Univ. (Grad. School of CSE, Hokkaido University)
2nd Author's Name Md. Al Mehedi Hasan  
2nd Author's Affiliation Research Institute for Electronic Science, Hokkaido University (RIES, Hokkaido Univ.)
3rd Author's Name Walker Peterson  
3rd Author's Affiliation Department of Chemistry, The University of Tokyo (Dept. of Chemistry, The Univ. of Tokyo)
4th Author's Name Koji Tabata  
4th Author's Affiliation Research Institute for Electronic Science, Hokkaido University (RIES, Hokkaido Univ.)
5th Author's Name Masahiro Sonoshita  
5th Author's Affiliation Institute for Genetic Medicine, Hokkaido University (Institute for Genetic Medicine, Hokkaido Univ.)
6th Author's Name Keisuke Goda  
6th Author's Affiliation Department of Chemistry, The University of Tokyo (Dept. of Chemistry, The Univ. of Tokyo)
7th Author's Name Tamiki Komatsuzaki  
7th Author's Affiliation Research Institute for Electronic Science, Hokkaido University (RIES, Hokkaido Univ.)
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 ()
21st Author's Name  
21st Author's Affiliation ()
22nd Author's Name  
22nd Author's Affiliation ()
23rd Author's Name  
23rd Author's Affiliation ()
24th Author's Name  
24th Author's Affiliation ()
25th Author's Name  
25th Author's Affiliation ()
26th Author's Name / /
26th Author's Affiliation ()
()
27th Author's Name / /
27th Author's Affiliation ()
()
28th Author's Name / /
28th Author's Affiliation ()
()
29th Author's Name / /
29th Author's Affiliation ()
()
30th Author's Name / /
30th Author's Affiliation ()
()
31st Author's Name / /
31st Author's Affiliation ()
()
32nd Author's Name / /
32nd Author's Affiliation ()
()
33rd Author's Name / /
33rd Author's Affiliation ()
()
34th Author's Name / /
34th Author's Affiliation ()
()
35th Author's Name / /
35th Author's Affiliation ()
()
36th Author's Name / /
36th Author's Affiliation ()
()
Speaker Author-1 
Date Time 2026-07-03 11:25:00 
Presentation Time 25 minutes 
Registration for IBISML 
Paper # NC2026-10, IBISML2026-10 
Volume (vol) vol.126 
Number (no) no.90(NC), no.91(IBISML) 
Page pp.38-45 
#Pages 8 
Date of Issue 2026-06-25 (NC, IBISML) 

[Return to Top Page]

[Return to IEICE Web Page]
The Institute of Electronics, Information and Communication Engineers (IEICE), Japan