pith. machine review for the scientific record. sign in

arxiv: 2603.29311 · v2 · submitted 2026-03-31 · ⚛️ physics.app-ph

Recognition: 2 theorem links

· Lean Theorem

Spoken Digit Recognition and Speaker Classification by Nonlinear Interfered Spin Wave-Based Physical Reservoir Computing

Sota Hikasa , Wataru Namiki , Daiki Nishioka , Maki Nishimura , Ryo Iguchi , Kazuya Terabe , Takashi Tsuchiya
Authors on Pith no claims yet

Pith reviewed 2026-05-13 23:56 UTC · model grok-4.3

classification ⚛️ physics.app-ph
keywords spin wavereservoir computingphysical reservoirspeech recognitionspeaker classificationspoken digit recognitionedge computingnonlinear dynamics
0
0 comments X

The pith

Nonlinear spin wave interference lets a physical reservoir computer classify speakers at 85.8 percent accuracy from raw audio signals alone.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper evaluates a spin-wave-based physical reservoir computer on spoken digit recognition and speaker classification. Four input configurations are tested to isolate the contribution of the physical reservoir versus conventional cochleagram preprocessing. The reservoir alone reaches 85.8 percent accuracy on speaker classification, close to the 90 percent obtained with cochleagram input. This shows the nonlinear dynamics of interfered spin waves can extract discriminative features directly from raw waveforms, reducing the need for costly frequency preprocessing in edge devices.

Core claim

The nonlinear interfered spin wave-based physical reservoir computing system achieves 85.8 percent accuracy for speaker classification when supplied only with raw speech signals, without cochleagram or other frequency-extraction preprocessing.

What carries the argument

The nonlinear interfered spin wave-based PRC, which maps raw time-series inputs into a high-dimensional state space via physical spin-wave interference and nonlinear dynamics for subsequent linear readout.

If this is right

  • Speaker classification becomes feasible on resource-limited hardware that cannot afford cochleagram computation.
  • Raw-waveform input to the reservoir lowers overall system latency and energy use for edge speech applications.
  • The same PRC architecture may generalize to other time-series tasks such as keyword spotting or anomaly detection in sensor streams.
  • Hardware implementations could bypass digital preprocessing stages entirely, simplifying the signal chain.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the raw-input performance holds, similar physical reservoirs could be evaluated on unprocessed video or vibration signals for broader sensory AI.
  • Integration of the spin-wave device with analog front-ends might enable always-on voice interfaces with sub-milliwatt power budgets.
  • The gap between cochleagram and raw accuracy on digit recognition versus speaker classification suggests task-specific tuning of reservoir parameters could close remaining shortfalls.

Load-bearing premise

The physical spin-wave system generates sufficient nonlinear transformations from raw audio to produce linearly separable features for speaker and digit classes.

What would settle it

Running the identical spin-wave reservoir on the same raw speech datasets but obtaining accuracy no higher than a linear baseline or below 70 percent on speaker classification would falsify the claim that the PRC extracts useful features without preprocessing.

read the original abstract

Recently, artificial-intelligence (AI) technologies have been increasingly utilized in a wide range of real-world applications. Speech recognition is one of these practical AI tasks and is regarded as a key application for edge AI systems. Consequently, speech recognition has been widely employed as a representative benchmark task for assessing the performance of physical reservoir computing (PRC). Although many PRCs have performed this task, the majority of them rely on the frequency-extraction preprocessing method, such as a cochleagram and mel-frequency cepstrum. Especially about the cochleagram, this method enables high-accuracy recognition; however, it requires a substantial computational cost for preprocessing and is unsuitable for edge computing, due to the limited resources. In this study, we employed a nonlinear interfered spin wave-based PRC, which demonstrated superior computational performance in mathematical tasks. Using this PRC, we evaluated the performance for two types of speech recognition, spoken digit recognition and speaker classification under four configurations: cochleagram-alone, interfered spin wave-based PRC with cochleagram, baseline without PRC, and interfered spin wave-based PRC alone to quantify the contributions of the cochleagram and of the interfered spin wave-based PRC for each task. As a result, although the cochleagram alone yielded accuracies around 90 % for both tasks, the accuracy reached 85.8 % for speaker classification when only the interfered spin wave-based PRC was used. These results indicate the potential of the proposed PRC to handle speech recognition tasks without cochleagram preprocessing.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript evaluates a nonlinear interfered spin-wave physical reservoir computing (PRC) system on spoken digit recognition and speaker classification. It reports results across four configurations (cochleagram alone, PRC with cochleagram, baseline without PRC, and PRC alone), claiming that the PRC alone reaches 85.8% accuracy on speaker classification and thereby demonstrates the potential to perform these tasks directly on raw audio without cochleagram preprocessing.

Significance. If the experimental claims hold after clarification of the input pipeline, the work would show that spin-wave PRC can extract discriminative features from unprocessed time-domain speech signals, offering a route to lower-preprocessing edge-AI speech systems. The multi-configuration design usefully isolates the reservoir's contribution relative to conventional preprocessing.

major comments (2)
  1. [Abstract] Abstract: the headline claim that the interfered spin-wave PRC alone achieves 85.8% speaker-classification accuracy rests on an underspecified input encoding. No details are supplied on sampling rate, amplitude normalization, windowing, filtering, or the mapping of the 1-D waveform onto the spin-wave device; any implicit frequency-selective operation would invalidate the assertion that the reservoir itself extracts the features from raw signals.
  2. [Results] Results/Methods (inferred from reported accuracies): the manuscript supplies no trial counts, error bars, statistical tests, or reservoir hyperparameters (device geometry, bias fields, readout training procedure). Without these, the numerical comparison between configurations cannot be verified and the quantification of each component's contribution remains unsupported.
minor comments (1)
  1. [Abstract] Abstract: the four configurations are listed but not enumerated clearly; a short numbered list would improve readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments. We agree with the need for more details on the input encoding and experimental parameters and have revised the manuscript to include them.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the headline claim that the interfered spin-wave PRC alone achieves 85.8% speaker-classification accuracy rests on an underspecified input encoding. No details are supplied on sampling rate, amplitude normalization, windowing, filtering, or the mapping of the 1-D waveform onto the spin-wave device; any implicit frequency-selective operation would invalidate the assertion that the reservoir itself extracts the features from raw signals.

    Authors: We thank the referee for highlighting this issue. We agree that the input encoding was not sufficiently detailed. In the revised manuscript, we will include a complete description of the sampling rate, amplitude normalization, windowing, filtering (none applied for the PRC-alone case), and the direct mapping of the 1-D time-domain waveform to the spin-wave device input. This will confirm that no frequency-selective operations are performed prior to the reservoir. revision: yes

  2. Referee: [Results] Results/Methods (inferred from reported accuracies): the manuscript supplies no trial counts, error bars, statistical tests, or reservoir hyperparameters (device geometry, bias fields, readout training procedure). Without these, the numerical comparison between configurations cannot be verified and the quantification of each component's contribution remains unsupported.

    Authors: We acknowledge that these details were missing from the original manuscript. In the revised version, we will report the number of trials, include error bars and statistical tests for the accuracy results, and provide all relevant reservoir hyperparameters such as device geometry, bias fields, and the readout training procedure. These additions will allow readers to verify the numerical comparisons and the contributions of each component. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical performance claims independent of any derivation chain

full rationale

The manuscript reports measured classification accuracies (e.g., 85.8 % speaker classification with interfered spin-wave PRC alone) against explicit baselines (cochleagram alone, combined, no-PRC). No equations, parameter fittings, or mathematical derivations appear that would reduce a claimed prediction to its own inputs by construction. The four configurations are presented as separate experimental runs; the 'PRC alone' result is not obtained by fitting to a subset and then relabeling the fit as a prediction. No self-citation load-bearing uniqueness theorem, ansatz smuggling, or renaming of known results is invoked. The paper is therefore self-contained against external benchmarks and receives the default non-circularity finding.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract contains no mathematical model, free parameters, axioms, or invented entities; the claim rests entirely on unreported experimental measurements.

pith-pipeline@v0.9.0 · 5597 in / 1001 out tokens · 36960 ms · 2026-05-13T23:56:32.250513+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

60 extracted references · 60 canonical work pages

  1. [1]

    Scalable reservoir computing on coherent linear photonic processor,

    M. Nakajima, K. Tanaka, and T. Hashimoto, “Scalable reservoir computing on coherent linear photonic processor,” Commun.,Phys. 4, 20 (2021). https://doi.org/10.1038/s42005-021-00519-1

    work page doi:10.1038/s42005-021-00519-1 2021

  2. [2]

    Embodiment of learning in electro- optical signal processors,

    M. Hermans, P. Antonik, M. Haelterman, and S. Massar, “Embodiment of learning in electro- optical signal processors,” Phys.,Rev.,Lett. 117, 128301 (2016). https://doi.org/10.1103/PhysRevLett.117.128301

    work page doi:10.1103/physrevlett.117.128301 2016

  3. [3]

    High-performance photonic reservoir computer based on a coherently driven passive cavity,

    Q. Vinckier, F. Duport, A. Smerieri, K. Vandoorne, P . Bienstman, M. Haelterman, and S. Masssar, “High-performance photonic reservoir computer based on a coherently driven passive cavity,” Optica 2, 438 (2015). https://doi.org/10.1364/OPTICA.2.000438

    work page doi:10.1364/optica.2.000438 2015

  4. [4]

    Fully analogue photonic reservoir computer,

    F. Duport, A. Smerieri, A. Akrout, M. Haelterman, and S. Massar, “Fully analogue photonic reservoir computer,” Sci.,Rep. 6, 22381 (2016). https://doi.org/10.1038/srep22381

    work page doi:10.1038/srep22381 2016

  5. [5]

    All-optical reservoir computing,

    F. Duport, B. Schneider, A. Smerieri, M. Haelterman, and S. Massar, “All-optical reservoir computing,” Opt.,Express 20, 22783 (2012). https://doi.org/10.1364/OE.20.022783

    work page doi:10.1364/oe.20.022783 2012

  6. [6]

    Optoelectronic reservoir computing,

    Y. Paquot, F. Duport, A. Smerieri, J. Dambre, B. Schrauwen, M. Haelterman, and S. Massar, “Optoelectronic reservoir computing,” Sci.,Rep.,2, 287 (2012). https://doi.org/10.1038/srep00287

    work page doi:10.1038/srep00287 2012

  7. [7]

    Edge-of-chaos learning achieved by ion-electron–coupled dynamics in an ion-gating reservoir,

    D. Nishioka, T. Tsuchiya, W. Namiki, M. Takayanagi, M. Imura, Y . Koide, T. Higuchi, and K. Terabe, “Edge-of-chaos learning achieved by ion-electron–coupled dynamics in an ion-gating reservoir,” Sci.,Adv. 8, eade1156 (2022). https://doi.org/10.1126/sciadv.ade1156

    work page doi:10.1126/sciadv.ade1156 2022

  8. [8]

    Few- and single-molecule reservoir computing experimentally demonstrated with surface-enhanced Raman scattering and ion gating,

    D. Nishioka, Y . Shingaya, T. Tsuchiya, T. Higuchi, and K. Terabe, “Few- and single-molecule reservoir computing experimentally demonstrated with surface-enhanced Raman scattering and ion gating,” Sci.,Adv. 10, eadk6438 (2024). https://doi.org/10.1126/sciadv.adk6438

    work page doi:10.1126/sciadv.adk6438 2024

  9. [9]

    Inverted input method for computing performance enhancement of the ion-gating reservoir,

    Y . Yamaguchi, D. Nishioka, W. Namiki, T. Tsuchiya, M. Imura, Y . Koide, T. Higuchi, and K. Terabe, “Inverted input method for computing performance enhancement of the ion-gating reservoir,” Appl. Phys.,Express 17, 024501 (2024). https://doi.org/10.35848/1882-0786/ad2906 19

    work page doi:10.35848/1882-0786/ad2906 2024

  10. [10]

    A high-performance deep reservoir computer experimentally demonstrated with ion-gating reservoirs,

    D. Nishioka, T. Tsuchiya, M. Imura, Y . Koide, T. Higuchi, and K. Terabe, “A high-performance deep reservoir computer experimentally demonstrated with ion-gating reservoirs,” Commun., Eng. 3, 81 (2024). https://doi.org/10.1038/s44172-024-00227-y

    work page doi:10.1038/s44172-024-00227-y 2024

  11. [11]

    Physical reservoir computing utilizing ion‐gating transistors operating in electric double layer and redox mechanisms,

    T. Tsuchiya, D. Nishioka, W. Namiki, and K. Terabe, “Physical reservoir computing utilizing ion‐gating transistors operating in electric double layer and redox mechanisms,” Adv.,Electron., Mater. 10, 2400625 (2024). https://doi.org/10.1002/aelm.202400625

    work page doi:10.1002/aelm.202400625 2024

  12. [12]

    A redox-based ion- gating reservoir, utilizing double reservoir states in drain and gate nonlinear response,

    T. Wada, D. Nishioka, W. Namiki, T. Tsuchiya, T. Higuchi, and K. Terabe “A redox-based ion- gating reservoir, utilizing double reservoir states in drain and gate nonlinear response,” Adv., Intell.,Syst. 5, 2300123 (2023). https://doi.org/10.1002/aisy.202300123

    work page doi:10.1002/aisy.202300123 2023

  13. [13]

    Redox-based ion- gating reservoir consisting of (104) oriented LiCoO2 film, assisted by physical masking,

    K. Shibata, D. Nishioka, W. Namiki, T. Tsuchiya, T. Higuchi, and K. Terabe, “Redox-based ion- gating reservoir consisting of (104) oriented LiCoO2 film, assisted by physical masking,” Sci., Rep. 13, 21060 (2023). https://doi.org/10.1038/s41598-023-48135-z

    work page doi:10.1038/s41598-023-48135-z 2023

  14. [14]

    Physical reservoir computing with graphene-based solid electric double layer transistor and the information processing capacity analysis,

    H. Kitano, D. Nishioka, K. Terabe, and T. Tsuchiya, “Physical reservoir computing with graphene-based solid electric double layer transistor and the information processing capacity analysis,” Appl.,Phys.,Express,18, 024501 (2025). https://doi.org/10.35848/1882-0786/adb19b

    work page doi:10.35848/1882-0786/adb19b 2025

  15. [15]

    Two orders of magnitude reduction in computational load achieved by ultrawideband responses of an ion- gating reservoir,

    D. Nishioka, H. Kitano, W. Namiki, S. Souma, K. Terabe, and T. Tsuchiya, “Two orders of magnitude reduction in computational load achieved by ultrawideband responses of an ion- gating reservoir,” ACS,Nano. 19, 42, 36896 (2025). https://doi.org/10.1021/acsnano.5c06174

    work page doi:10.1021/acsnano.5c06174 2025

  16. [16]

    Physical masking-induced enhancement of information processing capacity in a redox-type ion-gating reservoir,

    D. Nishioka, K. Shibata, W. Namiki, K. Terabe, and T. Tsuchiya, “Physical masking-induced enhancement of information processing capacity in a redox-type ion-gating reservoir,” Jpn.,J., Appl.,Phys.,64, 11SP25 (2025). https://doi.org/10.35848/1347-4065/ae1d84

    work page doi:10.35848/1347-4065/ae1d84 2025

  17. [17]

    Asymmetric transition of electrical resistance in an all-solid-state redox device with Fe3O4 and Li-ion electrolyte thin films for physical reservoir computing,

    W. Namiki, T. Tsuchiya, D. Nishioka, T. Higuchi, and K. Terabe, “Asymmetric transition of electrical resistance in an all-solid-state redox device with Fe3O4 and Li-ion electrolyte thin films for physical reservoir computing,” Jpn.,J.,Appl.,Phys. 63, 03SP13 (2024). https://doi.org/10.35848/1347-4065/ad1fb0

    work page doi:10.35848/1347-4065/ad1fb0 2024

  18. [18]

    Magnetization vector rotation reservoir computing operated by redox mechanism,

    W. Namiki, D. Nishioka, T. Tsuchiya, T. Higuchi, and K. Terabe, “Magnetization vector rotation reservoir computing operated by redox mechanism,” NanoLett. 24, 4383−4392 (2024). https://doi.org/10.1021/acs.nanolett.3c05029

    work page doi:10.1021/acs.nanolett.3c05029 2024

  19. [19]

    Reservoir computing with spin waves excited in a garnet film,

    R. Nakane, G. Tanaka, and A. Hirose, “Reservoir computing with spin waves excited in a garnet film,” IEEE,Access 6, 4462 (2018). https://doi.org/10.1109/ACCESS.2018.2794584

    work page doi:10.1109/access.2018.2794584 2018

  20. [20]

    Performance enhancement of a spin-wave-based reservoir computing system utilizing different physical conditions,

    R. Nakane, A. Hirose, and G. Tanaka, “Performance enhancement of a spin-wave-based reservoir computing system utilizing different physical conditions,” Phys.,Rev.,Appl.,19, 034047 (2023). https://doi.org/10.1103/PhysRevApplied.19.034047 20

    work page doi:10.1103/physrevapplied.19.034047 2023

  21. [21]

    Film-penetrating transducers applicable to on- chip reservoir computing with spin waves,

    J. Chen, R. Nakane, G. Tanaka, and A. Hirose, “Film-penetrating transducers applicable to on- chip reservoir computing with spin waves,” J.,Appl.,Phys. 132, 12 (2022). https://doi.org/10.1063/5.0102974

    work page doi:10.1063/5.0102974 2022

  22. [22]

    Nanoscale neural network using non-linear spin-wave interference,

    A. Papp, W. Porod, and G. Csaba, “Nanoscale neural network using non-linear spin-wave interference,” Nat.,Commun. 12, 6422 (2021). https://doi.org/10.1038/s41467-021-26711-z

    work page doi:10.1038/s41467-021-26711-z 2021

  23. [23]

    Characterization of nonlinear spin-wave interference by reservoir-computing metrics,

    A. Papp, G. Csaba, and W. Porod, “Characterization of nonlinear spin-wave interference by reservoir-computing metrics,” Appl.,Phys.,Lett. 119, 112403 (2021). https://doi.org/10.1063/5.0048982

    work page doi:10.1063/5.0048982 2021

  24. [24]

    Reservoir computing using a spin-wave delay-line active-ring resonator based on yttrium-iron-garnet film,

    S. Watt, and M. Kostylev, “Reservoir computing using a spin-wave delay-line active-ring resonator based on yttrium-iron-garnet film,” Phys.,Rev.,Applied 13, 034057 (2020). https://doi.org/10.1103/PhysRevApplied.13.034057

    work page doi:10.1103/physrevapplied.13.034057 2020

  25. [25]

    Implementing a magnonic reservoir computing model based on time-delay multiplexing,

    S. Watt, M. Kostylev, A. B. Ustinov, and B. A. Kalinikos, “Implementing a magnonic reservoir computing model based on time-delay multiplexing,” Phys.,Rev.,Applied 15, 064060 (2021). https://doi.org/10.1103/PhysRevApplied.15.064060

    work page doi:10.1103/physrevapplied.15.064060 2021

  26. [26]

    Spoken digit classification using a spin-wave delay-line active-ring reservoir computing,

    S. Watt, and M. Kostylev, “Spoken digit classification using a spin-wave delay-line active-ring reservoir computing,” arXiv 2005.12557 (2020). https://doi.org/10.48550/arXiv.2005.12557

    work page doi:10.48550/arxiv.2005.12557 2005

  27. [27]

    Experimental demonstration of high-performance physical reservoir computing with nonlinear interfered spin wave multidetection,

    W. Namiki, D. Nishioka, Y . Yamaguchi, T. Tsuchiya, Y . Higuchi, and K. Terabe, “Experimental demonstration of high-performance physical reservoir computing with nonlinear interfered spin wave multidetection,”,Adv.,Intell.,Syst.,5, 2300228 (2023). https://doi.org/10.1002/aisy.202300228

    work page doi:10.1002/aisy.202300228 2023

  28. [28]

    Fast physical reservoir computing, achieved with nonlinear interfered spin waves,

    W. Namiki, D. Nishioka, T. Tsuchiya, and K. Terabe, “Fast physical reservoir computing, achieved with nonlinear interfered spin waves,” Neuromorph.,Comput.,Eng.,4, 024015 (2024). https://doi.org/10.1088/2634-4386/ad561a

    work page doi:10.1088/2634-4386/ad561a 2024

  29. [29]

    Opto-magnonic reservoir computing coupling nonlinear interfered spin wave and visible light switching,

    W. Namiki, Y . Yamaguchi, D. Nishioka, T. Tsuchiya, and K. Terabe, “Opto-magnonic reservoir computing coupling nonlinear interfered spin wave and visible light switching,” Mater.,Today, Phys. 45, 10146 (2024). https://doi.org/10.1016/j.mtphys.2024.101465

    work page doi:10.1016/j.mtphys.2024.101465 2024

  30. [30]

    Iono– magnonic reservoir computing with chaotic spin wave interference manipulated by ion‐gating,

    W. Namiki, D. Nishioka, Y . Nomura, T. Tsuchiya, K. Yamamoto, and K. Terabe, “Iono– magnonic reservoir computing with chaotic spin wave interference manipulated by ion‐gating,” Adv.,Sci.,12, 2411777 (2025). https://doi.org/10.1002/advs.202411777

    work page doi:10.1002/advs.202411777 2025

  31. [31]

    Enhanced high dimensionality and the information processing capacity in interfered spin wave-based reservoir computing, achieved with eight detectors,

    S. Hikasa, W. Namiki, D. Nishioka, M. Nishimura, R. Iguchi, K. Terabe, and T. Tsuchiya, “Enhanced high dimensionality and the information processing capacity in interfered spin wave-based reservoir computing, achieved with eight detectors,” Adv.,Electron.,Mater.,E00780 (2026). https://doi.org/10.1002/aelm.202500780, 21

    work page doi:10.1002/aelm.202500780 2026

  32. [32]

    Spin-wave reservoir chips with short-term memory for high-speed estimation of external magnetic fields,

    S. Nagase, S. Nezu, and K. Sekiguchi ,“Spin-wave reservoir chips with short-term memory for high-speed estimation of external magnetic fields,”,Phys.,Rev.,Appl.,22, 024072 (2024). https://doi.org/10.1103/PhysRevApplied.22.024072

    work page doi:10.1103/physrevapplied.22.024072 2024

  33. [33]

    Task- adaptive physical reservoir computing,

    O. Lee, T. Wei, K. D. Stenning, J. C. Gartside, D. Prestwood, S. Seki, A. Aqeel, K. Karube, N. Kanazawa, Y . Taguchi, C. Back, Y . Tokura, W. R. Branford, and H. Kurebayashi, “Task- adaptive physical reservoir computing,” Nat.,Mater. 14, 79 (2023). https://doi.org/10.1038/s41563-023-01698-8

    work page doi:10.1038/s41563-023-01698-8 2023

  34. [34]

    Audio classification with skyrmion reservoirs,

    R. Msiska, J. Love, J. Mulkers, J. Leliaert, and K. Everschor-sitte, “Audio classification with skyrmion reservoirs,” Adv.,Intell.,Syst.,5, 2200388 (2023). https://doi.org/10.1002/aisy.202200388

    work page doi:10.1002/aisy.202200388 2023

  35. [35]

    Physical reservoir computing based on spin torque oscillator with forced synchronization,

    S. Tsunegi, T. Taniguchi, K. Nakajima, S. Miwa, K. Yakushiji, A. Fukushima, S. Yuasa, and H. Kubota, “Physical reservoir computing based on spin torque oscillator with forced synchronization,” Appl.,Phys.,Lett. 114, 164101 (2019). https://doi.org/10.1063/1.5081797

    work page doi:10.1063/1.5081797 2019

  36. [36]

    Neuromorphic computing with nanoscale spintronic oscillator,

    J. Torrejon, M. Riou, F. A. Araujo, S. Tsunegi, G. Khalsa, D. Querlioz, P. Bortolotti, V. Cros, K. Yakushiji, A. Fukushima, H. Kubota, S. Yuasa, M. D. Stiles, and J. Grollier, “Neuromorphic computing with nanoscale spintronic oscillator,” Nature.,547, 428-432 (2017). https://doi.org/10.1038/nature23011

    work page doi:10.1038/nature23011 2017

  37. [37]

    Memristor-based spiking neural networks: cooperative development of neural network architecture/algorithms and memristors,

    H. Peng, L. Gan, and X. Guo, “Memristor-based spiking neural networks: cooperative development of neural network architecture/algorithms and memristors,” Chip,3, 100093 (2024). https://doi.org/10.1016/j.chip.2024.100093

    work page doi:10.1016/j.chip.2024.100093 2024

  38. [38]

    C. Du, F. Cai, M.A. Zidan, W. Ma, S. H. Lee, and W. D. Lu,” Reservoir computing using dynamic memristors for temporal information processing,”,Nat.,Commun. 8, 2204 (2017). https://doi.org/10.1038/s41467-017-02337-y

    work page doi:10.1038/s41467-017-02337-y 2017

  39. [39]

    In,materia,reservoir computing with a fully memristive architecture based on self- organizing nanowire networks,

    G. Milano, G. Pedretti, K. Montano, S. Ricci, S. Hashemkhani, L. Boarino, D. Ielmini, and C. Ricciardi, “In,materia,reservoir computing with a fully memristive architecture based on self- organizing nanowire networks,” Nat.,Mater. 21, 195-202 (2022). https://doi.org/10.1038/s41563-021-01099-9

    work page doi:10.1038/s41563-021-01099-9 2022

  40. [40]

    Speech recognition through physical reservoir computing with neuromorphic nanowire networks,

    G. Milano, M. Agliuzza, N. Leo, and C. Ricciardi, “Speech recognition through physical reservoir computing with neuromorphic nanowire networks,” Proceedings,of,the,2022, International,Joint,Conference,on,Neural,Networks,(IJCNN),18-23 (2022). https://doi.org/10.1109/IJCNN55064.2022.9892078

    work page doi:10.1109/ijcnn55064.2022.9892078 2022

  41. [41]

    Simulation platform for pattern recognition based on reservoir computing with memristor networks,

    G. Tanaka, and R. Nakane, “Simulation platform for pattern recognition based on reservoir computing with memristor networks,” Sci.,Rep.,12, 9868 (2022). https://doi.org/10.1038/s41598-022-13687-z 22

    work page doi:10.1038/s41598-022-13687-z 2022

  42. [42]

    Dynamic memristor-based reservoir computing for high-efficiency temporal signal processing,

    Y . Zhong, J. Tang, X. Li, B. Gao, H. Qian, and H. Wu, “Dynamic memristor-based reservoir computing for high-efficiency temporal signal processing,” Nat.,Commun. 12, 408 (2021). https://doi.org/10.1038/s41467-020-20692-1

    work page doi:10.1038/s41467-020-20692-1 2021

  43. [43]

    Information processing via physical soft body,

    K. Nakajima, H. Hauser, T. Li, and R. Pfeifer,,“Information processing via physical soft body,”, Sci.,Rep. 5, 10487 (2015). https://doi.org/10.1038/srep10487

    work page doi:10.1038/srep10487 2015

  44. [44]

    Hardware-friendly implementation of physical reservoir computing with CMOS-based time-domain analog spiking neurons,

    N. Kimura, C. Duran, Z. Byambadorj, R. Nakane, and T. Iizuka, “Hardware-friendly implementation of physical reservoir computing with CMOS-based time-domain analog spiking neurons,” Neuromorph.,Comput.,Eng.,5, 024015 (2025). https://doi.org/10.1088/2634- 4386/addb6d

    work page doi:10.1088/2634- 2025

  45. [45]

    In-materio reservoir computing in a sulfonated polyaniline network,

    Y . Usami, B. v. d. Ven, D. G. Mathew, T. Chen, T. Kotooka, Y .Kawashima, Y . Tanaka, Y . Otsuka, H. Ohoyama, H. Tamukoh, H.Tanaka, W. G. v. d. Wiel, and T. Matsumoto “In-materio reservoir computing in a sulfonated polyaniline network,”,Adv.,Mater.,33, 2102688 (2021). https://doi.org/10.1002/adma.202102688

    work page doi:10.1002/adma.202102688 2021

  46. [46]

    Intrinsic Disordered Network in Multiferroic YMnO3 Single Crystals for In-Materio Physical Reservoir Computing Through Tuneable Domain-Wall Structure,

    M. Xu, K. Furuta, A. Karacali, Y . Umezaki, A. Zaman, Y . Usami, H. Tanaka, and Y . Horibe, “Intrinsic Disordered Network in Multiferroic YMnO3 Single Crystals for In-Materio Physical Reservoir Computing Through Tuneable Domain-Wall Structure,” Small e06397 (2025). https://doi.org/10.1002/smll.202506397

    work page doi:10.1002/smll.202506397 2025

  47. [47]

    Hopf physical reservoir computer for reconfigurable sound recognition,

    M. R. E. U. Shougat, X. Li, S. Shao, K. McGarvey, and E. Perkins, “Hopf physical reservoir computer for reconfigurable sound recognition,” Sci.,Rep.,13, 8719 (2023). https://doi.org/10.1038/s41598-023-35760-x

    work page doi:10.1038/s41598-023-35760-x 2023

  48. [48]

    Reservoir Computing System With HZO/Si FeFETs in Parallel Configuration: Experimental Demonstration of Speech Classification,

    E. Nako, K. Toprasertpong, R. Nakane, M. Takenaka, and S. Takagi, “Reservoir Computing System With HZO/Si FeFETs in Parallel Configuration: Experimental Demonstration of Speech Classification,” 2023,IEEE,Trans.,Electron,Devices ,70, 5657-64 (2023). https://doi.org/10.1109/TED.2023.3318870

    work page doi:10.1109/ted.2023.3318870 2023

  49. [49]

    Information processing using a single dynamical node as complex system,

    L. Appeltant, M.C. Soriano, G.V . Sande, J. Dawnckaert, S. Massar, J. Dambre, B. Schrauwen, C. R. Mirasso, and I. Fischer, “Information processing using a single dynamical node as complex system,” Nat.,Commun. 2, 468 (2011). https://doi.org/10.1038/ncomms1476

    work page doi:10.1038/ncomms1476 2011

  50. [50]

    Speech recognition with cochlea-inspired in-sensor computing,

    P. H. Beoletto, G. Milano, C. Ricciardi, F. Bosia, and A. S. Gliozzi, “Speech recognition with cochlea-inspired in-sensor computing,” Adv.,Intell.,Syst.,8, 2500526 (2026). https://doi.org/10.1002/aisy.202500526

    work page doi:10.1002/aisy.202500526 2026

  51. [51]

    Role of non-linear data processing on speech recognition task in the framework of reservoir computing,

    F. A. Araujo, M. Riou, J. Torrejon, S. Tsunegi, D. Querlioz, K. Yakushiji, A. Fukushima, H. Kubota, S. Yuasa, M. D. Stiles, and J. Grollier, “Role of non-linear data processing on speech recognition task in the framework of reservoir computing,” Sci.,Rep.,10, 328 (2020). https://doi.org/10.1038/s41598-019-56991-x 23

    work page doi:10.1038/s41598-019-56991-x 2020

  52. [52]

    Recent advances in physical reservoir computing: A review,

    G. Tanaka, T. Yamane, J. B. Heroux, R. Nakane, N. Kanazawa, S. Takeda, H. Numata, D. Nakano, and A. Hirose, “Recent advances in physical reservoir computing: A review,”,Neural, Networks 115, 100-123 (2019). https://doi.org/10.1016/j.neunet.2019.03.005

    work page doi:10.1016/j.neunet.2019.03.005 2019

  53. [53]

    The Echo State

    H. Jaeger, “The Echo State” approach to analysing and training recurrent neural networks-with an erratum note,” German,Nat.,Res.,Center,for,Inf.,Technol.,GMD,Technol. 148, 13 (2010)

    work page 2010

  54. [54]

    Freely scalable and reconfigurable optical hardware for deep learning,

    L. Bernstein, A. Sludds, R. Hamerly, V . Sze, J. Emer, and D. Englund, “Freely scalable and reconfigurable optical hardware for deep learning,” Sci.,Rep. 11, 3144 (2021). https://doi.org/10.1038/s41598-021-82543-3

    work page doi:10.1038/s41598-021-82543-3 2021

  55. [55]

    Emerging opportunities and challenges for the future of reservoir computing,

    M. Yan, C. Huang, P. Bienstman, P. Tino, W. Lin, and J. Sun, “Emerging opportunities and challenges for the future of reservoir computing,”,Nat.,Commun. 15, 2056 (2024). https://doi.org/10.1038/s41467-024-45187-1

    work page doi:10.1038/s41467-024-45187-1 2056

  56. [56]

    A computational model of filtering, detection and compression in the cochlea,

    R. Lyon,,“A computational model of filtering, detection and compression in the cochlea,” ICASSP’82.,IEEE,Int.,Conf.,on,Acoustics ,Speech,and,Signal,Processing,7,,3-5 (1982). https://doi.org/10.1109/ICASSP.1982.1171644

    work page doi:10.1109/icassp.1982.1171644 1982

  57. [57]

    The impact of the MIT-BIH Arrhythmia Database,

    GB. Moody, RG. Mark. “The impact of the MIT-BIH Arrhythmia Database,” IEEE,Eng,in,Med, and,Biol 20, 45-50 (2001). http://ecg.mit.edu/george/publications/mitdb-embs-2001.pdf

    work page 2001

  58. [58]

    Introduction to spin wave computing,

    A. Mahmou, F. Ciubotaru, F. Vanderveken, A. V . Chumak, S. Hamdioui, C. Adelmann, and S. Cotofana, “Introduction to spin wave computing,” J.,Appl.,Phys. 128, 161101 (2020). https://doi.org/10.1063/5.0019328

    work page doi:10.1063/5.0019328 2020

  59. [59]

    Review on spintronics: Principles and device applications , volume =

    A. Hirohata, K. Yamada, Y . Nakatani, I. Prejbeanu, B. Dieny, P. Pirro, and B. Hillebrands, “Review on Spintronics: Principles and Device Applications,” Journal,of,Magnetism,and, Magnetic,Materials 50, 166711 (2020). https://doi.org/10.1016/j.jmmm.2020.166711

    work page doi:10.1016/j.jmmm.2020.166711 2020

  60. [60]

    Liberman,et al.,,TI,46-Word,LDC93S9.,Web,Download.,Philadelphia:,Linguistic,Data, Consortium,(1993),(,https://doi.org/10.35111/zx7a-fw03 )

    M. Liberman,et al.,,TI,46-Word,LDC93S9.,Web,Download.,Philadelphia:,Linguistic,Data, Consortium,(1993),(,https://doi.org/10.35111/zx7a-fw03 )

    work page doi:10.35111/zx7a-fw03 1993