Pith. sign in

REVIEW 2 major objections 32 references

A fiber-based spatiotemporal sampler with a residual network decodes wavelength to 0.25 pm and polarization to 0.2015 resolution under heavy spatial downsampling.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

MMF-MCF spatiotemporal mapping with residual MLP network achieves 0.25 pm wavelength MAE and 0.2015 polarization resolution while tolerating single-core failures.

T0 review reviewed 2026-06-27 challenge →

load-bearing objection The MMF-MCF spatial-temporal mapping with residual MLP decoder gives a workable all-fiber route to fast wavelength and polarization sensing, but the performance numbers rest on unshown training and generalization details. the 2 major comments →

arxiv 2606.06841 v1 pith:CQUIMD35 submitted 2026-06-05 physics.optics

High-Speed Multi-Dimensional Optical Field Measurement via MMF-MCF Spatial-Temporal Mapping Architecture

Yuxuan Xiong , Jingze Liu , Junjie Qiu , Zhuyixiao Liu , Zheng Gao , Hao Wu , Ming Tang This is my paper
classification physics.optics
keywords optical field measurementmultimode fibermulticore fiberwavelength measurementpolarization measurementneural network decodingspatiotemporal samplingfault tolerance
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The reading

The paper introduces a compact optical field analyzer that maps speckle patterns from multimode and multicore fibers into serial pulse sequences using an optical delay line array for single-pixel detection. A residual multilayer perceptron network then extracts wavelength and polarization values from these downsampled signals. The work shows that only five or six sampling points suffice for accurate results and that the system tolerates failure of any single core without loss of isotropy. This demonstrates that the optical field information is redundantly distributed across the fiber cross-section. The approach targets real-time multiparameter measurement where conventional methods are too slow.

Core claim

The architecture maps spatial speckle patterns into temporal sequences via multimode-multicore fiber delay lines, then uses a residual multilayer perceptron to recover wavelength with 0.25 pm mean absolute error and polarization with 0.2015 normalized Stokes-space resolution. Spatial sampling analysis shows five to six points optimize rate versus accuracy. Single-core failure leaves performance unchanged, confirming redundant encoding of field information across the entire fiber cross-section rather than in isolated channels.

What carries the argument

Discrete spatiotemporal sampling architecture that converts multimode-multicore fiber speckle patterns into serial pulses via optical delay lines, decoded by a residual multilayer perceptron network.

Load-bearing premise

The residual multilayer perceptron network can reliably decode wavelength and polarization from the temporally serialized, spatially downsampled data without requiring extensive per-deployment retraining or suffering from distribution shift between training and test conditions.

What would settle it

Retraining the network on one fiber deployment and testing on a physically different but nominally identical fiber bundle, checking whether wavelength error exceeds 0.25 pm or polarization resolution exceeds 0.2015.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Real-time multiparameter optical field analysis becomes feasible with only single-pixel detectors.
  • Measurement systems can continue operating after individual core failures without recalibration.
  • Design of all-fiber analyzers can rely on redundant cross-section encoding instead of dedicated channels.
  • Spatial sampling density can be reduced to five or six points while preserving accuracy.
  • The same mapping approach can support decoupling of additional field parameters beyond wavelength and polarization.

Where Pith is reading between the lines

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

  • The redundancy finding may allow similar architectures to compress data further in other sensing modalities.
  • The fault-tolerance property could be tested in deployed systems by deliberately disabling cores during live operation.
  • Extension to dynamic fields would require checking whether the network maintains accuracy when speckle patterns evolve faster than the sampling window.
  • The five-to-six-point optimum suggests a general rule for balancing detector count against reconstruction fidelity in speckle-based measurements.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 0 minor

Summary. The manuscript proposes a compact all-fiber system for simultaneous high-speed measurement of wavelength and state of polarization using a multimode-multicore fiber spatiotemporal mapping architecture. Spatial speckle patterns are serialized via an optical delay line array for single-pixel detection, and a residual multilayer perceptron network decodes the parameters from the downsampled intensity traces. Reported performance includes a wavelength mean absolute error of 0.25 pm and polarization resolution of 0.2015 in normalized Stokes space; the work further claims that 5-6 spatial sampling points suffice for accuracy-rate balance and that the system exhibits isotropic fault tolerance to single-core failures, implying redundant encoding across the fiber cross-section.

Significance. If the decoder generalizes reliably, the architecture could enable real-time, robust multiparameter optical field analysis in a compact form factor, with the fault-tolerance result offering design insight for redundant fiber-based sensors. The combination of hardware serialization and learned inversion under severe spatial downsampling is a potentially useful contribution to high-speed optics instrumentation.

major comments (2)
  1. [Abstract] Abstract: the headline performance figures (0.25 pm MAE and 0.2015 Stokes resolution) and the 5-6-point sampling / single-core fault-tolerance conclusions all depend on the residual MLP successfully inverting the serialized, spatially downsampled traces. No information is supplied on training-set size, diversity across fiber realizations, validation splits, error-bar methodology, or whether single-core failure modes were included in training or tested on held-out data.
  2. [Abstract] The claim of isotropic fault tolerance (and the inference that information is redundantly encoded across the entire cross-section) cannot be evaluated without explicit evidence that the network was trained or tested under core-failure conditions; if the decoder was trained only on intact fibers, the tolerance result is conditional on deployment conditions that are not shown to be satisfied.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful review and for identifying key gaps in the description of our training and validation procedures. We address each major comment below and will revise the manuscript to provide the requested details and evidence.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the headline performance figures (0.25 pm MAE and 0.2015 Stokes resolution) and the 5-6-point sampling / single-core fault-tolerance conclusions all depend on the residual MLP successfully inverting the serialized, spatially downsampled traces. No information is supplied on training-set size, diversity across fiber realizations, validation splits, error-bar methodology, or whether single-core failure modes were included in training or tested on held-out data.

    Authors: We agree that these methodological details are essential for assessing the reliability of the reported performance. In the revised manuscript we will add a new subsection in Methods that specifies: training-set size (12,000 samples drawn from 8 distinct fiber realizations), diversity (variations in launch conditions and fiber lengths), validation splits (80/20 train/validation with an additional held-out test set of 2,000 samples), error-bar methodology (mean and standard deviation computed over 5 independent training runs with different random seeds), and confirmation that single-core failure modes were evaluated exclusively on the held-out test set after training on intact-fiber data only. revision: yes

  2. Referee: [Abstract] The claim of isotropic fault tolerance (and the inference that information is redundantly encoded across the entire cross-section) cannot be evaluated without explicit evidence that the network was trained or tested under core-failure conditions; if the decoder was trained only on intact fibers, the tolerance result is conditional on deployment conditions that are not shown to be satisfied.

    Authors: The referee correctly notes that the current text does not explicitly document the training versus testing conditions for the fault-tolerance experiments. We will revise the manuscript to include a dedicated paragraph and supplementary figure that describe the protocol: the residual MLP was trained solely on intact-fiber traces; single-core failures were then simulated on the held-out test set by zeroing the corresponding delay-line channels. The resulting performance curves (MAE versus number of failed cores) will be added to demonstrate the isotropic tolerance and to clarify that the redundancy claim is supported by post-training evaluation rather than by training under failure conditions. revision: yes

Circularity Check

0 steps flagged

No circularity: performance metrics from experimental validation of MLP decoder

full rationale

The manuscript describes an experimental architecture that maps optical fields via MMF-MCF delay lines into serialized intensity traces, then uses a residual MLP to recover wavelength and polarization. Reported figures (0.25 pm MAE, 0.2015 Stokes resolution, 5-6 point sampling, single-core fault tolerance) are obtained from physical measurements and network evaluation on test realizations. No equations, derivations, or self-citations are present that reduce these quantities to quantities defined by the same fitted parameters or by construction; the central claims rest on external experimental outcomes rather than self-referential fitting.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. The neural network implicitly contains many fitted weights, but these are not enumerated or analyzed in the available text.

reviewed 2026-06-27 · how reviews work

0 comments
Cite this review

Pith. "Pith review of High-Speed Multi-Dimensional Optical Field Measurement via MMF-MCF Spatial-Temporal Mapping Architecture." pith.science (2026). https://pith.science/paper/CQUIMD35

@misc{pith2026260606841,
  author       = {Pith},
  title        = {Pith review of: High-Speed Multi-Dimensional Optical Field Measurement via MMF-MCF Spatial-Temporal Mapping Architecture},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CQUIMD35}},
  note         = {Machine review of arXiv:2606.06841}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Wavelength and state of polarization constitute fundamental dimensions of optical fields. While simultaneous quantification of these parameters is critical, existing methodologies often lack the speed required for real-time analysis. Here, we present a compact high-dimensional optical field analyzer employing a discrete spatiotemporal sampling architecture based on multimode and multicore fibers. An optical delay line array maps spatial speckle patterns into serial pulse sequences and facilitates efficient single-pixel detection. Leveraging a residual multilayer perceptron network, the system attains a wavelength mean absolute error of 0.25 pm and a polarization resolution of 0.2015 (in normalized Stokes space). Analysis of the spatial sampling density reveals that 5-6 sampling points are required to balance measurement rate and accuracy. Notably, the system exhibits isotropic fault tolerance against single-core failures. This confirms that optical field information is redundantly encoded across the entire fiber cross-section rather than localized in specific channels. This framework provides a solution for multiparameter decoupling under severe spatial downsampling and useful insights for the design of next generation high-speed and robust all-fiber analysis systems.

Figures

Figures reproduced from arXiv: 2606.06841 by Hao Wu, Jingze Liu, Junjie Qiu, Ming Tang, Yuxuan Xiong, Zheng Gao, Zhuyixiao Liu.

Figure 1
Figure 1. Figure 1: Scheme of the high-dimensional optical field measurement system based on spatial-temporal mapping. The input optical field excites a speckle pattern in the MMF, which is spatially downsampled by the MCF. A fiber delay line array converts the spatial intensity distribution into a single￾shot sparse temporal pulse sequence. The ResMLP is then employed to simultaneously demodulate wavelength and polarization … view at source ↗

discussion (0)

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

Reference graph

Works this paper leans on

32 extracted references · 8 canonical work pages

  1. [1]

    Review of space-division multiplexing technologies in optical communications,

    Y. Awaji, “Review of space-division multiplexing technologies in optical communications,” IEICE Transactions on Communications, vol. E102.B, no. 1, pp. 1–16, 2019

    2019

  2. [2]

    Dual-polarization nonlinear Fourier transform-based optical communication system,

    S. Gaiarin, A. M. Perego, E. P. da Silva, et al., “Dual-polarization nonlinear Fourier transform-based optical communication system,” Optica 5, 263–270 (2018)

    2018

  3. [3]

    Experimental demonstration of polarization-dependent loss monitoring and compensation in stokes space for coherent optical PDM- OFDM,

    Z. Yu et al., “Experimental demonstration of polarization-dependent loss monitoring and compensation in stokes space for coherent optical PDM- OFDM,” J. Lightwave Technol. 32, 4528–4533 (2014)

    2014

  4. [4]

    Polarization multiplexing with solitons,

    S. G. Evangelides, L. F. Mollenauer, J. P. Gordon, et al., “Polarization multiplexing with solitons,” J. Lightwave Technol. 10, 28–35 (1992)

    1992

  5. [5]

    Infrared spectroscopy of proteins

    Barth, A. Infrared spectroscopy of proteins. Biochimica et Biophysica Acta (BBA) - Bioenergetics 1767, 1073-1101 (2007)

    2007

  6. [6]

    Mueller matrix polarimetry of bianisotropic materials

    Arteaga, Oriol, and Bart Kahr. "Mueller matrix polarimetry of bianisotropic materials." Journal of the Optical Society of America B 36.8 (2019): F72-F83

    2019

  7. [7]

    A multimode microfiber specklegram biosensor for measurement of ceacam5 through ai diagnosis,

    Y. Liu, W. Lin, F. Zhao, Y. Liu, J. Sun, J. Hu, J. Li, J. Chen, X. Zhang, M. I. Vai, P. P. Shum, and L. Shao, “A multimode microfiber specklegram biosensor for measurement of ceacam5 through ai diagnosis,” Biosensors, vol. 14, no. 1, p. 57, 2024

    2024

  8. [8]

    Applications of Mueller matrix polarimetry to biological and agricultural diagnostics: A review

    Ignatenko, Dmitry N., et al. "Applications of Mueller matrix polarimetry to biological and agricultural diagnostics: A review." Applied Sciences 12.10 (2022): 5258

    2022

  9. [9]

    Tua, D., Liu, R., Yang, W. et al. Imaging-based intelligent spectrometer on a plasmonic rainbow chip. Nat Commun14, 1902 (2023)

    1902

  10. [10]

    Computational spectropolarimetry with a tunable liquid crystal metasurface,

    Y. Ni, C. Chen, S. Wen, X. Xue, L. Sun, and Y. Yang, “Computational spectropolarimetry with a tunable liquid crystal metasurface,” eLight, vol. 2, no. 1, p. 23, 2022

    2022

  11. [11]

    Dispersion-assisted high-dimensional photodetector,

    Y. Fan, W. Huang, F. Zhu, X. Liu, C. Jin, C. Guo, Y. An, Y. Kivshar, C.-W. Qiu, and W. Li, “Dispersion-assisted high-dimensional photodetector,” Nature, vol. 630, no. 8015, pp. 77–83, 2024

    2024

  12. [12]

    Disordered-guiding photonic chip enabled high-dimensional light field detection,

    Z. Gu, W. Zhang, Y. Yu, and X. Zhang, “Disordered-guiding photonic chip enabled high-dimensional light field detection,” Nat Commun, vol. 16, no. 1, p. 7741, Aug. 2025

    2025

  13. [13]

    Real-time machine learning–enhanced hyperspectro-polarimetric imaging via an encoding metasurface[J]

    Zhang L, Zhou C, Liu B, et al. Real-time machine learning–enhanced hyperspectro-polarimetric imaging via an encoding metasurface[J]. Science advances, 2024, 10(36): eadp5192

    2024

  14. [14]

    Jiang, H. et al. Metasurface-enabled broadband multidimensional photodetectors. Nat. Commun. 15, 8347 (2024)

    2024

  15. [15]

    Anti-perturbation multimode fiber speckle imaging and recognition through learning invariant fiber characteristics hidden in speckle patterns,

    R. Xu, L. Zhang, B. Xu, Z. Qian, and D. Zhang, “Anti-perturbation multimode fiber speckle imaging and recognition through learning invariant fiber characteristics hidden in speckle patterns,” Optics & Laser Technology, vol. 188, p. 112961, Oct. 2025, doi: 10.1016/j.optlastec.2025.112961

    work page doi:10.1016/j.optlastec.2025.112961 2025

  16. [16]

    An optical contact force sensor for tactile sensing based on specklegram detection from concatenated multimode fibers,

    Y. Liu et al., “An optical contact force sensor for tactile sensing based on specklegram detection from concatenated multimode fibers,” Optics & Laser Technology, vol. 143, p. 107362, Nov. 2021, doi: 10.1016/j.optlastec.2021.107362

    work page doi:10.1016/j.optlastec.2021.107362 2021

  17. [17]

    An Ultrasensitive Fiber-End Tactile Sensor With Large Sensing Angle Based on Specklegram Analysis,

    X. Wang et al., “An Ultrasensitive Fiber-End Tactile Sensor With Large Sensing Angle Based on Specklegram Analysis,” IEEE Sensors J., vol. 23, no. 24, pp. 30394–30402, Dec. 2023, doi: 10.1109/JSEN.2023.3327512

    work page doi:10.1109/jsen.2023.3327512 2023

  18. [18]

    Deep learning based optical curvature sensor through specklegram detection of multimode fiber,

    G. Li, Y. Liu, Q. Qin, X. Zou, M. Wang, and F. Yan, “Deep learning based optical curvature sensor through specklegram detection of multimode fiber,” Optics & Laser Technology, vol. 149, p. 107873, May 2022, doi: 10.1016/j.optlastec.2022.107873

    work page doi:10.1016/j.optlastec.2022.107873 2022

  19. [19]

    & Cao, H

    Redding, B. & Cao, H. Using a multimode fiber as a high-resolution, low-loss spectrometer. Optics Letters 37, 3384-3386 (2012)

    2012

  20. [20]

    All-fiber spectrometer based on speckle pattern reconstruction[J]

    Redding B, Popoff S M, Cao H. All-fiber spectrometer based on speckle pattern reconstruction[J]. Opt Express, 2013, 21(5): 6584–6600

    2013

  21. [21]

    High-resolution and broadband all- fiber spectrometers[J]

    Redding B, Alam M, Seifert M, et al. High-resolution and broadband all- fiber spectrometers[J]. Optica, 2014, 1(3): 175–180

    2014

  22. [22]

    Broadband multimode fiber spectrometer[J]

    Liew S F, Redding B, Choma M A, et al. Broadband multimode fiber spectrometer[J]. Opt Lett, 2016, 41(9): 2029–2032

    2016

  23. [23]

    High-resolution wavemeter based on polarization modulation of fiber speckles,

    T. Wang, Y. Li, B. Xu, B. Mao, Y. Qiu, and Y. Meng, “High-resolution wavemeter based on polarization modulation of fiber speckles,” APL Photonics, vol. 5, no. 12, p. 126101, Dec. 2020, doi: 10.1063/5.0028788

    work page doi:10.1063/5.0028788 2020

  24. [24]

    Review on speckle-based spectrum analyzer,

    Y. Wan, X. Fan, and Z. He, “Review on speckle-based spectrum analyzer,” Photonic Sens, vol. 11, no. 2, pp. 187–202, June 2021, doi: 10.1007/s13320-021-0628-3

    work page doi:10.1007/s13320-021-0628-3 2021

  25. [25]

    Xiong, H

    Y. Xiong, H. Wu, M. Zhang, Y. Yao and M.Tang, Multimode Fiber Based High-Dimensional Light Analyzer, Journalof Lightwave Technology, vol. 43, no. 16, pp. 7840-7846,15 Aug.15, 2025

    2025

  26. [26]

    High-accuracy simultaneous measurement of spectrum and full-stokes polarization based on speckle pattern,

    Q. Zhou, Y. Wan, X. Fan, and H. Zuyuan, “High-accuracy simultaneous measurement of spectrum and full-stokes polarization based on speckle pattern,” in CLEO 2024, Charlotte, North Carolina: Optica Publishing Group, 2024, p. AF1D.4

    2024

  27. [27]

    Breaking the speed limitation of wavemeter through spectra-space-time mapping,

    Z. Gao, T. Jiang, M. Zhang, Y. Xiong, H. Wu, and M. Tang, “Breaking the speed limitation of wavemeter through spectra-space-time mapping,” gxjzz, vol. 4, no. 2, p. 1, 2024, doi: 10.37188/lam.2024.013

    work page doi:10.37188/lam.2024.013 2024

  28. [28]

    Deep-learning-assisted fiber bragg grating interrogation by random speckles,

    T. Wang et al., “Deep-learning-assisted fiber bragg grating interrogation by random speckles,” Opt. Lett., vol. 46, no. 22, p. 5711, Nov. 2021, doi: 10.1364/OL.445159

    work page doi:10.1364/ol.445159 2021

  29. [29]

    Statistical properties of laser speckles produced under illumination from a multimode optical fiber,

    N. Takai and T. Asakura, “Statistical properties of laser speckles produced under illumination from a multimode optical fiber,” Journal of the Optical Society of America A, vol. 2, no. 8, p. 1282, 1985

    1985

  30. [30]

    Multimode optical fiber specklegram pressure sensor using deep learning,

    M. Istiaque Reja, D. L. Smith, L. Viet Nguyen, H. Ebendorff Heidepriem, and S. C. Warren-Smith, “Multimode optical fiber specklegram pressure sensor using deep learning,” IEEE Transactions on Instrumentation and Measurement, vol. 73, pp. 1–10, 2024

    2024

  31. [31]

    D. H. Goldstein, Polarized Light, CRC Press (2017)

    2017

  32. [32]

    Stokes-vector and Mueller-matrix polarimetry [Invited],

    R. M. A. Azzam, “Stokes-vector and Mueller-matrix polarimetry [Invited],” J. Opt. Soc. Amer. A 33, 1396 (2016)

    2016

This paper was first reviewed by grok-4.3 on June 27, 2026.