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 2026-03-04 16:15
FedExPred: Design and Analysis of an Explainable Federated EKF-LSTM Localisation Framework
Lisa Johnson-Davies, Yuto Lim, Yasuo Tan (JAIST) IN2025-74
Abstract (in Japanese) (See Japanese page) 
(in English) Federated vehicle localisation must cope with heterogeneous environments while avoiding centralisation of sensitive trajectory data. Under non-IID clients, however, naive aggregation can amplify brittle client-specific cues and degrade cross-environment generalisation. This work is design-oriented: it presents the structure of an explainable federated EKF–LSTM localisation framework and highlights the tuning levers that govern its behaviour. Each client runs an Extended Kalman Filter (EKF) baseline and trains a lightweight LSTM to predict residual pose corrections. To explain the learned correction (rather than generic motion dynamics), we compute Integrated Gradients (IG) using an EKF-scoped baseline and compress per-feature attributions into a compact explanation profile. On the server, Explain–Then–Weight (ETW) treats aggregation as a configurable scoring rule that blends local validation error with explanation-profile similarity; its behaviour is controlled by weights and an optional damping factor, alongside choices such as similarity metric and attribution summarisation. Using a nuScenes-derived four-client split with leave-one-client-out evaluation, we report preliminary results intended to illustrate heterogeneity sensitivity: EKF+LSTM tends to lower centralised error, while ETW can yield modest gains over FedAvg in some held-out environments. These observations motivate a design-and-tuning roadmap for controlled scenario splits and robustness tests (e.g., GNSS dropout) in future work.
Keyword (in Japanese) (See Japanese page) 
(in English) Autonomous Vehicles / Localisation / Federated Learning / Explainable AI / Integrated Gradients / EKF / LSTM /  
Reference Info. IEICE Tech. Rep., vol. 125, no. 386, IN2025-74, pp. 67-72, March 2026.
Paper # IN2025-74 
Date of Issue 2026-02-25 (IN) 
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 IN2025-74

Conference Information
Committee IN NS  
Conference Date 2026-03-04 - 2026-03-06 
Place (in Japanese) (See Japanese page) 
Place (in English) Okinawa-Ken Shichoson Jichi Kaikan 
Topics (in Japanese) (See Japanese page) 
Topics (in English) General 
Paper Information
Registration To IN 
Conference Code 2026-03-IN-NS 
Language English 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) FedExPred: Design and Analysis of an Explainable Federated EKF-LSTM Localisation Framework 
Sub Title (in English)  
Keyword(1) Autonomous Vehicles  
Keyword(2) Localisation  
Keyword(3) Federated Learning  
Keyword(4) Explainable AI  
Keyword(5) Integrated Gradients  
Keyword(6) EKF  
Keyword(7) LSTM  
Keyword(8)  
1st Author's Name Lisa Johnson-Davies  
1st Author's Affiliation Japan Advanced Institute of Science and Technology (JAIST)
2nd Author's Name Yuto Lim  
2nd Author's Affiliation Japan Advanced Institute of Science and Technology (JAIST)
3rd Author's Name Yasuo Tan  
3rd Author's Affiliation Japan Advanced Institute of Science and Technology (JAIST)
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 ()
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-2 
Date Time 2026-03-04 16:15:00 
Presentation Time 25 minutes 
Registration for IN 
Paper # IN2025-74 
Volume (vol) vol.125 
Number (no) no.386 
Page pp.67-72 
#Pages 6 
Date of Issue 2026-02-25 (IN) 

[Return to Top Page]

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