Preprocessing Methods for Memristive Reservoir Computing for Image Recognition

Rishona Daniels; Duna Wattad; Ronny Ronen; David Saad; Shahar Kvatinsky

doi:10.1109/MetroXRAINE66377.2025.11340063

Preprocessing Methods for Memristive Reservoir Computing for Image Recognition

Rishona Daniels
, Duna Wattad
, Ronny Ronen
, David Saad
, Shahar Kvatinsky

Chemical Engineering and Applied Chemistry, College of Engineering and Physical Sciences Aston University Birmingham B4 7ET United Kingdom
Technion - Israel Institute of Technology,Viterbi Faculty of Electrical and Computer Engineering,Haifa,Israel

Research output: Chapter in Book/Published conference output › Conference publication

Abstract

Reservoir computing (RC) has attracted attention as an efficient recurrent neural network architecture due to its simplified training, requiring only its last perceptron readout layer to be trained. When implemented with memristors, RC systems benefit from their dynamic properties, which make them ideal for reservoir construction. However, achieving high performance in memristor-based RC remains challenging, as it critically depends on the input preprocessing method and reservoir size. Despite growing interest, a comprehensive evaluation that quantifies the impact of these factors is still lacking. This paper systematically compares various preprocessing methods for memristive RC systems, assessing their effects on accuracy and energy consumption. We also propose a parity-based preprocessing method that improves accuracy by 2-6% while requiring only a modest increase in device count compared to other methods. Our findings highlight the importance of informed preprocessing strategies to improve the efficiency and scalability of memristive RC systems.

Original language	English
Title of host publication	Conference Proceedings - 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering, MetroXRAINE 2025
Publisher	IEEE
Pages	1164-1169
Number of pages	6
ISBN (Electronic)	9798331502799
DOIs	https://doi.org/10.1109/MetroXRAINE66377.2025.11340063
Publication status	Published - 23 Jan 2026
Event	2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE) - Ancona, Italy Duration: 22 Oct 2025 → 24 Oct 2025

Publication series

Name	MetroXRAINE Proceedings
Publisher	IEEE

Conference

Conference	2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE)
Period	22/10/25 → 24/10/25

Bibliographical note

Copyright © 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Funding

This research has been partially funded by the European Union's Horizon 2020 Research And Innovation Programme FET-Open NEU-Chip under grant agreement No. 964877 and funded by the European Union (ERC, Real-Database-PIM, 101157452). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council Executive Agency. Neither the European Union nor the granting authority can be held responsible for them.

Funders	Funder number
European Commission
Horizon 2020 Framework Programme	964877
European Research Council	101157452

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

memristors
neuromorphic computing
reservoir computing

Access to Document

10.1109/MetroXRAINE66377.2025.11340063

https://arxiv.org/abs/2506.05588Licence: CC BY-NC-ND 4.0

Cite this

Daniels, R., Wattad, D., Ronen, R., Saad, D., & Kvatinsky, S. (2026). Preprocessing Methods for Memristive Reservoir Computing for Image Recognition. In Conference Proceedings - 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering, MetroXRAINE 2025 (pp. 1164-1169). (MetroXRAINE Proceedings). IEEE. https://doi.org/10.1109/MetroXRAINE66377.2025.11340063

@inproceedings{8c304c89decd46df80f1d7daa0be6bcf,

title = "Preprocessing Methods for Memristive Reservoir Computing for Image Recognition",

abstract = "Reservoir computing (RC) has attracted attention as an efficient recurrent neural network architecture due to its simplified training, requiring only its last perceptron readout layer to be trained. When implemented with memristors, RC systems benefit from their dynamic properties, which make them ideal for reservoir construction. However, achieving high performance in memristor-based RC remains challenging, as it critically depends on the input preprocessing method and reservoir size. Despite growing interest, a comprehensive evaluation that quantifies the impact of these factors is still lacking. This paper systematically compares various preprocessing methods for memristive RC systems, assessing their effects on accuracy and energy consumption. We also propose a parity-based preprocessing method that improves accuracy by 2-6\% while requiring only a modest increase in device count compared to other methods. Our findings highlight the importance of informed preprocessing strategies to improve the efficiency and scalability of memristive RC systems.",

keywords = "memristors, neuromorphic computing, reservoir computing",

author = "Rishona Daniels and Duna Wattad and Ronny Ronen and David Saad and Shahar Kvatinsky",

note = "Copyright {\textcopyright} 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.; 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE) ; Conference date: 22-10-2025 Through 24-10-2025",

year = "2026",

month = jan,

day = "23",

doi = "10.1109/MetroXRAINE66377.2025.11340063",

language = "English",

series = "MetroXRAINE Proceedings",

publisher = "IEEE",

pages = "1164--1169",

booktitle = "Conference Proceedings - 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering, MetroXRAINE 2025",

address = "United States",

}

Daniels, R, Wattad, D, Ronen, R, Saad, D & Kvatinsky, S 2026, Preprocessing Methods for Memristive Reservoir Computing for Image Recognition. in Conference Proceedings - 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering, MetroXRAINE 2025. MetroXRAINE Proceedings, IEEE, pp. 1164-1169, 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE), 22/10/25. https://doi.org/10.1109/MetroXRAINE66377.2025.11340063

Preprocessing Methods for Memristive Reservoir Computing for Image Recognition. / Daniels, Rishona; Wattad, Duna; Ronen, Ronny et al.
Conference Proceedings - 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering, MetroXRAINE 2025. IEEE, 2026. p. 1164-1169 (MetroXRAINE Proceedings).

Research output: Chapter in Book/Published conference output › Conference publication

TY - GEN

T1 - Preprocessing Methods for Memristive Reservoir Computing for Image Recognition

AU - Daniels, Rishona

AU - Wattad, Duna

AU - Ronen, Ronny

AU - Saad, David

AU - Kvatinsky, Shahar

N1 - Copyright © 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

PY - 2026/1/23

Y1 - 2026/1/23

N2 - Reservoir computing (RC) has attracted attention as an efficient recurrent neural network architecture due to its simplified training, requiring only its last perceptron readout layer to be trained. When implemented with memristors, RC systems benefit from their dynamic properties, which make them ideal for reservoir construction. However, achieving high performance in memristor-based RC remains challenging, as it critically depends on the input preprocessing method and reservoir size. Despite growing interest, a comprehensive evaluation that quantifies the impact of these factors is still lacking. This paper systematically compares various preprocessing methods for memristive RC systems, assessing their effects on accuracy and energy consumption. We also propose a parity-based preprocessing method that improves accuracy by 2-6% while requiring only a modest increase in device count compared to other methods. Our findings highlight the importance of informed preprocessing strategies to improve the efficiency and scalability of memristive RC systems.

AB - Reservoir computing (RC) has attracted attention as an efficient recurrent neural network architecture due to its simplified training, requiring only its last perceptron readout layer to be trained. When implemented with memristors, RC systems benefit from their dynamic properties, which make them ideal for reservoir construction. However, achieving high performance in memristor-based RC remains challenging, as it critically depends on the input preprocessing method and reservoir size. Despite growing interest, a comprehensive evaluation that quantifies the impact of these factors is still lacking. This paper systematically compares various preprocessing methods for memristive RC systems, assessing their effects on accuracy and energy consumption. We also propose a parity-based preprocessing method that improves accuracy by 2-6% while requiring only a modest increase in device count compared to other methods. Our findings highlight the importance of informed preprocessing strategies to improve the efficiency and scalability of memristive RC systems.

KW - memristors

KW - neuromorphic computing

KW - reservoir computing

UR - https://ieeexplore.ieee.org/document/11340063

UR - https://www.scopus.com/pages/publications/105033214497

U2 - 10.1109/MetroXRAINE66377.2025.11340063

DO - 10.1109/MetroXRAINE66377.2025.11340063

M3 - Conference publication

T3 - MetroXRAINE Proceedings

SP - 1164

EP - 1169

BT - Conference Proceedings - 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering, MetroXRAINE 2025

PB - IEEE

T2 - 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE)

Y2 - 22 October 2025 through 24 October 2025

ER -

Daniels R, Wattad D, Ronen R, Saad D, Kvatinsky S. Preprocessing Methods for Memristive Reservoir Computing for Image Recognition. In Conference Proceedings - 2025 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering, MetroXRAINE 2025. IEEE. 2026. p. 1164-1169. (MetroXRAINE Proceedings). Epub 2025 Oct 22. doi: 10.1109/MetroXRAINE66377.2025.11340063

Preprocessing Methods for Memristive Reservoir Computing for Image Recognition

Abstract

Publication series

Conference

Bibliographical note

Funding

UN SDGs

Keywords

Access to Document

Other files and links

Embargoed Document

Fingerprint

Cite this