All-optical Edge Computing for Speckle Sensing Interrogation

Catarina S. Monteiro; Joana M. Teixeira; Nuno A. Silva; Pedro A. S. Jorge; Tiago D. Ferreira; Tom\'as Lopes

arxiv: 2605.25805 · v1 · pith:JXHCKUB7new · submitted 2026-05-25 · ⚛️ physics.optics

All-optical Edge Computing for Speckle Sensing Interrogation

Tom\'as Lopes , Joana M. Teixeira , Tiago D. Ferreira , Catarina S. Monteiro , Pedro A. S. Jorge , Nuno A. Silva This is my paper

Pith reviewed 2026-06-29 20:33 UTC · model grok-4.3

classification ⚛️ physics.optics

keywords speckle sensingall-optical computingedge computingdigital micromirror devicesignal separationoptical fiber sensorevolutionary optimizationreal-time sensing

0 comments

The pith

A programmable optical mask trained only from detector feedback separates multiple concurrent speckle signals in real time.

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

The paper demonstrates that speckle sensing can skip camera capture and digitization by moving the computation into the optical path. A digital micromirror device projects the speckle pattern and applies masks whose settings are adjusted in place by an evolutionary optimizer that sees only the final detector output. In a fiber sensor perturbed by several piezoelectric actuators at once, the trained masks extract individual signals with more than 4 dB gain and less than -10 dB crosstalk. Because no imaging hardware is required, the measurement rate is set solely by the photodetector speed. This removes the latency and bandwidth limits that have kept speckle sensors from operating in real time.

Core claim

The manuscript presents an all-optical edge-computing platform that performs task-specific computation directly in the optical domain for speckle-based sensors. Speckle patterns are projected onto a digital micromirror device used as a programmable optical layer whose parameters are trained in situ by evolutionary optimization driven solely by detector feedback. In a multi-point fiber sensing experiment, multiple piezoelectric actuators simultaneously perturb the fiber; the optimized masks decouple the concurrent signals, delivering target enhancement above 4 dB and crosstalk suppression below -10 dB while operating at bandwidths limited only by the photodetector.

What carries the argument

A digital micromirror device acting as an in-situ-trained programmable optical mask layer whose settings are found by evolutionary optimization from single-detector feedback alone.

If this is right

Sensing bandwidth becomes limited only by photodetector response rather than camera frame rate or digitization speed.
Multiple independent measurands encoded in one speckle field can be read out simultaneously without electronic image processing.
The optical layer can be retrained on-site whenever the sensing task changes, using only the existing detector.
Latency between perturbation and signal output drops to the propagation time through the optics and detector electronics.

Where Pith is reading between the lines

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

The same mask-training approach could be applied to other high-dimensional optical fields where camera readout is the current bottleneck.
Replacing the DMD with a faster spatial light modulator would allow the method to track perturbations on microsecond or nanosecond scales.
Because training requires no camera or labeled data, the platform could be deployed in environments where imaging hardware is impractical.

Load-bearing premise

Evolutionary optimization driven only by detector readings will converge to masks that continue to separate signals when any combination of the concurrent perturbations occurs.

What would settle it

Run the trained masks on simultaneous actuator drives whose exact combination was never presented during optimization and observe whether crosstalk rises above -10 dB or target enhancement falls below 4 dB.

read the original abstract

Speckle-based sensing exploits the rich environmental information of its high-dimensional spatial intensity patterns. However, the requirement for camera-based acquisition and subsequent electronic digitization introduces significant latency and bandwidth bottlenecks that forbid real-time operation and higher temporal resolutions. Aiming to bypass this imaging processing pipeline, this manuscript presents an optically reconfigurable edge-computing platform for speckle-based sensors that performs task-specific computation directly in the optical domain. This is achieved by projecting output speckle patterns onto a digital micromirror device, using it as a programmable optical layer whose parameters are trained in situ using an evolutionary optimization strategy solely from detector feedback. We demonstrate the concept with a multi-point optical fiber sensing task, where multiple piezoelectric actuators simultaneously perturb the fiber, modifying the speckle pattern. Optimizing a set of masks to decouple these concurrent signals, the system successfully achieves real-time signal separation, achieving a target signal enhancement exceeding 4 dB while suppressing crosstalk leakage below -10 dB. Operating with bandwidths limited only by the photodetector, this approach paves the way for real-time and ultrafast optical sensing via an all-optical edge computing solution.

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.

Desk Editor's Note private letter to a colleague

DMD masks trained from photodetector feedback separate multi-actuator speckle signals at detector speeds, but the abstract leaves generalization and validation details unaddressed.

read the letter

The main point is that this paper replaces camera-based speckle readout with a DMD that applies learned masks, where the masks are tuned by a genetic algorithm that sees only the output voltage from a single photodetector. They test it on a fiber perturbed by several piezoelectric actuators at once and report over 4 dB signal enhancement with crosstalk below -10 dB, all running at photodetector bandwidth.

The new element is the in-situ evolutionary training driven solely by detector feedback for this multi-point separation task. That setup directly targets the latency and bandwidth limits of camera digitization in speckle sensing.

The demonstration itself is presented cleanly in the abstract, with concrete performance numbers tied to the optical preprocessing step. The framing around bypassing electronic processing is practical and matches a known constraint in the field.

The soft spot is the missing experimental detail. No mention of trial counts, error bars, or camera ground-truth comparison appears in the abstract. The stress-test concern holds: detector-only fitness can reward masks that separate the specific training drives without guaranteeing separation for arbitrary simultaneous perturbations, especially if actuator-induced speckle changes are linearly dependent. Without held-out tests or post-training validation, the reported metrics rest on that unstated assumption.

This is for researchers in fiber-optic sensing who need higher temporal resolution than cameras allow. The concept is concrete enough that it deserves a serious referee to examine the methods section and any generalization checks.

Referee Report

2 major / 0 minor

Summary. The manuscript presents an all-optical edge-computing platform for speckle-based sensing interrogation. A DMD is used as a programmable optical layer whose masks are trained in situ via evolutionary optimization using only integrated photodetector intensity as fitness. For a multi-point optical fiber sensing task with concurrent piezoelectric actuator perturbations, the optimized masks enable real-time signal separation, with reported performance of >4 dB target signal enhancement and <-10 dB crosstalk leakage at photodetector-limited bandwidths.

Significance. If the performance metrics are reproducible and generalize, the approach could enable real-time, high-bandwidth speckle sensing without camera latency or bandwidth limits, advancing all-optical computation for sensing. The in-situ detector-only training is a practical strength for deployment. However, the current description supplies no experimental details, error bars, trial counts, or validation protocols, so the significance cannot yet be assessed.

major comments (2)

[Section 3] Section 3: The GA is stated to operate solely on detector voltage with no camera feedback during training and no post-selection mentioned. The manuscript provides no information on training-set size, number of generations, actuator drive dimensionality, or explicit tests for generalization to held-out simultaneous perturbations. Without these, it is unclear whether the reported >4 dB enhancement and <-10 dB crosstalk reflect true separability or merely local optima that separate the training drives but fail for linearly dependent speckle changes across the operating regime.
[Methods] Methods: The abstract and text report quantitative metrics (>4 dB enhancement, <-10 dB crosstalk) with no accompanying experimental details such as number of independent trials, error bars, specific GA hyperparameters, or direct comparison of the separated signals against camera-based ground truth. These omissions make it impossible to determine whether the stated performance supports the central claim of reliable, real-time signal separation.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments, which highlight areas where additional detail will improve the manuscript. We address each major comment below and will incorporate the requested information in a revised version.

read point-by-point responses

Referee: [Section 3] Section 3: The GA is stated to operate solely on detector voltage with no camera feedback during training and no post-selection mentioned. The manuscript provides no information on training-set size, number of generations, actuator drive dimensionality, or explicit tests for generalization to held-out simultaneous perturbations. Without these, it is unclear whether the reported >4 dB enhancement and <-10 dB crosstalk reflect true separability or merely local optima that separate the training drives but fail for linearly dependent speckle changes across the operating regime.

Authors: We agree that Section 3 requires expanded description of the GA procedure and validation. In the revision we will add the training-set size, number of generations, actuator drive dimensionality, and explicit results from held-out simultaneous perturbations. These additions will demonstrate that the reported metrics arise from separability that generalizes beyond the training drives rather than from local optima. revision: yes
Referee: [Methods] Methods: The abstract and text report quantitative metrics (>4 dB enhancement, <-10 dB crosstalk) with no accompanying experimental details such as number of independent trials, error bars, specific GA hyperparameters, or direct comparison of the separated signals against camera-based ground truth. These omissions make it impossible to determine whether the stated performance supports the central claim of reliable, real-time signal separation.

Authors: We concur that the Methods section must supply the missing experimental details for reproducibility. The revised manuscript will include the number of independent trials, error bars on all reported metrics, the specific GA hyperparameters, and a direct comparison of the optically separated signals against camera-based ground truth. These additions will allow readers to assess the reliability of the real-time separation performance. revision: yes

Circularity Check

0 steps flagged

No derivation chain present; results are direct experimental measurements

full rationale

The manuscript describes an experimental demonstration of an all-optical edge-computing platform. Masks on a DMD are optimized via genetic algorithm using only integrated photodetector intensity as the fitness function, and the reported performance (>4 dB enhancement, <-10 dB crosstalk) is obtained by direct measurement of detector output under actuator drives. No equations, first-principles derivations, fitted parameters renamed as predictions, or uniqueness theorems are invoked. The central claim is therefore an empirical observation rather than a mathematical reduction that could be circular by construction. No self-citation load-bearing steps appear in the provided text.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Based solely on the abstract, the work rests on standard domain assumptions in optics and evolutionary optimization with no explicit free parameters, new entities, or ad-hoc axioms introduced.

axioms (1)

domain assumption Speckle patterns encode rich environmental information from fiber perturbations
Stated as the foundation for speckle-based sensing in the opening sentence.

pith-pipeline@v0.9.1-grok · 5747 in / 1208 out tokens · 13695 ms · 2026-06-29T20:33:43.201424+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

26 extracted references · 15 canonical work pages

[1]

Light: Science and Applications6(2017) https://doi.org/10.1038/ lsa.2017.35

Makris, K.G., Brandst¨ otter, A., Ambichl, P., Musslimani, Z.H., Rotter, S.: Wave propagation through disordered media without backscatter- ing and intensity variations. Light: Science and Applications6(2017) https://doi.org/10.1038/ lsa.2017.35

2017
[2]

Optics Express27, 28494 (2019) https:// doi.org/10.1364/oe.27.028494

Murray, M.J., Davis, A., Kirkendall, C., Redding, B.: Speckle-based strain sensing in multimode fiber. Optics Express27, 28494 (2019) https:// doi.org/10.1364/oe.27.028494

work page doi:10.1364/oe.27.028494 2019
[3]

Optics Express28, 20830 (2020) https://doi.org/10.1364/oe.384423

Bennett, A., Beiderman, Y., Agdarov, S., Beider- man, Y., Hendel, R., Straussman, B., Zalevsky, Z.: Monitoring of vital bio-signs by analysis of speckle patterns in a fabric-integrated multimode opti- cal fiber sensor. Optics Express28, 20830 (2020) https://doi.org/10.1364/oe.384423

work page doi:10.1364/oe.384423 2020
[4]

Optics Communications573(2024) https://doi

Zhu, Y., Mi, Z., Zhu, Y., Ke, C., Rong, L., Shi, Y.: Optical information hiding based on speckle encoding with dual-multiplexing interferometry. Optics Communications573(2024) https://doi. org/10.1016/j.optcom.2024.131022

work page doi:10.1016/j.optcom.2024.131022 2024
[5]

AI agents vs

Wang, X., Tao, Y., Yang, F., Zhang, Y.: An effective compressive computational ghost imag- ing with hybrid speckle pattern. Optics Commu- nications454(2020) https://doi.org/10.1016/j. optcom.2019.124470

work page doi:10.1016/j 2020
[6]

In: 2023 9th International Workshop on Advances in Sensors and Interfaces (IWASI), pp

K¨ uhne, J., Magno, M., Benini, L.: A fast and accu- rate optical flow camera for resource-constrained edge applications. In: 2023 9th International Workshop on Advances in Sensors and Interfaces (IWASI), pp. 143–148. IEEE, ??? (2023). https: //doi.org/10.1109/IWASI58316.2023.10164626 . https://ieeexplore.ieee.org/document/10164626/

work page doi:10.1109/iwasi58316.2023.10164626 2023
[7]

IEEE Transactions on Pattern Analysis and Machine Intelligence44, 154–180 (2022) https://doi.org/ 10.1109/TPAMI.2020.3008413

Gallego, G., Delbruck, T., Orchard, G., Bar- tolozzi, C., Taba, B., Censi, A., Leutenegger, S., Davison, A.J., Conradt, J., Daniilidis, K., Scara- muzza, D.: Event-based vision: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence44, 154–180 (2022) https://doi.org/ 10.1109/TPAMI.2020.3008413

work page doi:10.1109/tpami.2020.3008413 2022
[8]

arXiv preprint arXiv:2509.22094 (2025)

Lopes, T., Teixeira, J., Rocha, V.V., Fer- reira, T.D., Monteiro, C.S., Jorge, P.A., Silva, N.A.: Event-based speckle interrogation for high- bandwidth multi-point optical fiber sensing. arXiv preprint arXiv:2509.22094 (2025)

work page arXiv 2025
[9]

Light: Science and Applications8(2019) https://doi.org/10.1038/ s41377-019-0223-1

Luo, Y., Mengu, D., Yardimci, N.T., Rivenson, Y., Veli, M., Jarrahi, M., Ozcan, A.: Design of task-specific optical systems using broadband diffractive neural networks. Light: Science and Applications8(2019) https://doi.org/10.1038/ s41377-019-0223-1

2019
[10]

Springer (2024)

Fu, T., Zhang, J., Sun, R., Huang, Y., Xu, W., Yang, S., Zhu, Z., Chen, H.: Optical neural net- works: progress and challenges. Springer (2024). https://doi.org/10.1038/s41377-024-01590-3

work page doi:10.1038/s41377-024-01590-3 2024
[11]

Communications Physics8(1), 68 (2025)

Skalli, A., Goldmann, M., Haghighi, N., Reitzen- stein, S., Lott, J.A., Brunner, D.: Annealing- inspired training of an optical neural network with ternary weights. Communications Physics8(1), 68 (2025)

2025
[12]

Computer Communi- cations194, 213–225 (2022) https://doi.org/10

Caiazza, C., Giordano, S., Luconi, V., Vecchio, A.: Edge computing vs centralized cloud: Impact of communication latency on the energy consump- tion of lte terminal nodes. Computer Communi- cations194, 213–225 (2022) https://doi.org/10. 1016/j.comcom.2022.07.026

2022
[13]

IEEE Sensors Journal25, 31545–31562 (2025) https: //doi.org/10.1109/JSEN.2025.3584460

Delaney, E., Brophy, T., Ward, E., Collins, F., Jones, E., Deegan, B., Glavin, M.: Evaluating event-based vision sensing in rain and fog. IEEE Sensors Journal25, 31545–31562 (2025) https: //doi.org/10.1109/JSEN.2025.3584460

work page doi:10.1109/jsen.2025.3584460 2025
[14]

Nature Photonics18, 1076–1082 (2024) https: //doi.org/10.1038/s41566-024-01494-z 7

Yildirim, M., Dinc, N.U., Oguz, I., Psaltis, D., Moser, C.: Nonlinear processing with linear optics. Nature Photonics18, 1076–1082 (2024) https: //doi.org/10.1038/s41566-024-01494-z 7

work page doi:10.1038/s41566-024-01494-z 2024
[15]

eLight4(2024) https://doi.org/10.1186/ s43593-024-00067-5

Hu, J., Liao, K., Din¸ c, N.U., Gigli, C., Bai, B., Gan, T., Li, X., Chen, H., Yang, X., Li, Y., I¸ sıl, Rahman, M.S.S., Li, J., Hu, X., Jarrahi, M., Psaltis, D., Ozcan, A.: Subwavelength imaging using a solid-immersion diffractive optical pro- cessor. eLight4(2024) https://doi.org/10.1186/ s43593-024-00067-5

2024
[16]

Light: Science and Applications7(2018) https://doi

Rahmani, B., Loterie, D., Konstantinou, G., Psaltis, D., Moser, C.: Multimode optical fiber transmission with a deep learning network. Light: Science and Applications7(2018) https://doi. org/10.1038/s41377-018-0074-1

work page doi:10.1038/s41377-018-0074-1 2018
[17]

Optics Express33, 36133 (2025) https://doi.org/ 10.1364/oe.571745

Lan, R., Li, Y., Shen, H., Lu, Y., Hu, H., Zheng, L., Zong, Q., Zang, Y., Zhang, Z.: All-optical color image encryption using multimode fiber speckles and diffractive deep neural networks. Optics Express33, 36133 (2025) https://doi.org/ 10.1364/oe.571745

work page doi:10.1364/oe.571745 2025
[18]

Optics and Laser Technology160(2023) https://doi.org/10.1016/ j.optlastec.2022.109062

Cai, Y., Liu, Y., Li, G., Qin, Q., Pang, L., Ren, W., Wei, J., Wang, M.: Reflective tactile sensor assisted by multimode fiber-based optical cou- pler and fiber specklegram. Optics and Laser Technology160(2023) https://doi.org/10.1016/ j.optlastec.2022.109062

work page arXiv 2023
[19]

Nanomaterials13(2023) https: //doi.org/10.3390/nano13040768

Gu, L., Gao, H., Hu, H.: Demonstration of a learning-empowered fiber specklegram sensor based on focused ion beam milling for refrac- tive index sensing. Nanomaterials13(2023) https: //doi.org/10.3390/nano13040768

work page doi:10.3390/nano13040768 2023
[20]

IEEE Sensors Journal (2025)

Uyar, F., Aldemir, A.S., Kartaloglu, T., Ozbay, E., Ozdur, I.: A compliant mandrel-based fiber optic hydrophone for underwater acoustic sensing. IEEE Sensors Journal (2025)

2025
[21]

IEEE Transactions on Power Delivery25(4), 2526–2534 (2010)

Lima, S.E., Fraz˜ ao, O., Farias, R.G., Ara´ ujo, F.M., Ferreira, L.A., Santos, J.L., Miranda, V.: Mandrel-based fiber-optic sensors for acoustic detection of partial discharges—a proof of con- cept. IEEE Transactions on Power Delivery25(4), 2526–2534 (2010)

2010
[22]

IUCrJ10(5), 568–578 (2023)

Wang, C., Florin, E., Chang, H.-Y., Thayer, J., Yoon, C.H.: Specklenn: a unified embedding for real-time speckle pattern classification in x-ray single-particle imaging with limited labeled exam- ples. IUCrJ10(5), 568–578 (2023)

2023
[23]

Physics Letters A468, 128736 (2023)

Castilho, V., Balthazar, W., Silva, L., Penna, T., Huguenin, J.: Machine learning classification of speckle patterns for roughness measurements. Physics Letters A468, 128736 (2023)

2023
[24]

IEEE Sensors Journal23(18), 20937– 20950 (2023)

Newaz, A., Faruque, M.O., Al Mahmud, R., Sagor, R.H., Khan, M.Z.M.: Machine-learning- enabled multimode fiber specklegram sensors: a review. IEEE Sensors Journal23(18), 20937– 20950 (2023)

2023
[25]

Optics express 23(9), 11898–11911 (2015)

Dr´ emeau, A., Liutkus, A., Martina, D., Katz, O., Sch¨ ulke, C., Krzakala, F., Gigan, S., Daudet, L.: Reference-less measurement of the transmission matrix of a highly scattering material using a dmd and phase retrieval techniques. Optics express 23(9), 11898–11911 (2015)

2015
[26]

SurfaceCodeQuantumCommunication,

Popoff, S.M., Lerosey, G., Carminati, R., Fink, M., Boccara, A.C., Gigan, S.: Measuring the transmission matrix in optics: An approach to the study and control of light propagation in disordered media. Physical Review Letters104 (2010) https://doi.org/10.1103/PhysRevLett.104. 100601 8

work page doi:10.1103/physrevlett.104 2010

[1] [1]

Light: Science and Applications6(2017) https://doi.org/10.1038/ lsa.2017.35

Makris, K.G., Brandst¨ otter, A., Ambichl, P., Musslimani, Z.H., Rotter, S.: Wave propagation through disordered media without backscatter- ing and intensity variations. Light: Science and Applications6(2017) https://doi.org/10.1038/ lsa.2017.35

2017

[2] [2]

Optics Express27, 28494 (2019) https:// doi.org/10.1364/oe.27.028494

Murray, M.J., Davis, A., Kirkendall, C., Redding, B.: Speckle-based strain sensing in multimode fiber. Optics Express27, 28494 (2019) https:// doi.org/10.1364/oe.27.028494

work page doi:10.1364/oe.27.028494 2019

[3] [3]

Optics Express28, 20830 (2020) https://doi.org/10.1364/oe.384423

Bennett, A., Beiderman, Y., Agdarov, S., Beider- man, Y., Hendel, R., Straussman, B., Zalevsky, Z.: Monitoring of vital bio-signs by analysis of speckle patterns in a fabric-integrated multimode opti- cal fiber sensor. Optics Express28, 20830 (2020) https://doi.org/10.1364/oe.384423

work page doi:10.1364/oe.384423 2020

[4] [4]

Optics Communications573(2024) https://doi

Zhu, Y., Mi, Z., Zhu, Y., Ke, C., Rong, L., Shi, Y.: Optical information hiding based on speckle encoding with dual-multiplexing interferometry. Optics Communications573(2024) https://doi. org/10.1016/j.optcom.2024.131022

work page doi:10.1016/j.optcom.2024.131022 2024

[5] [5]

AI agents vs

Wang, X., Tao, Y., Yang, F., Zhang, Y.: An effective compressive computational ghost imag- ing with hybrid speckle pattern. Optics Commu- nications454(2020) https://doi.org/10.1016/j. optcom.2019.124470

work page doi:10.1016/j 2020

[6] [6]

In: 2023 9th International Workshop on Advances in Sensors and Interfaces (IWASI), pp

K¨ uhne, J., Magno, M., Benini, L.: A fast and accu- rate optical flow camera for resource-constrained edge applications. In: 2023 9th International Workshop on Advances in Sensors and Interfaces (IWASI), pp. 143–148. IEEE, ??? (2023). https: //doi.org/10.1109/IWASI58316.2023.10164626 . https://ieeexplore.ieee.org/document/10164626/

work page doi:10.1109/iwasi58316.2023.10164626 2023

[7] [7]

IEEE Transactions on Pattern Analysis and Machine Intelligence44, 154–180 (2022) https://doi.org/ 10.1109/TPAMI.2020.3008413

Gallego, G., Delbruck, T., Orchard, G., Bar- tolozzi, C., Taba, B., Censi, A., Leutenegger, S., Davison, A.J., Conradt, J., Daniilidis, K., Scara- muzza, D.: Event-based vision: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence44, 154–180 (2022) https://doi.org/ 10.1109/TPAMI.2020.3008413

work page doi:10.1109/tpami.2020.3008413 2022

[8] [8]

arXiv preprint arXiv:2509.22094 (2025)

Lopes, T., Teixeira, J., Rocha, V.V., Fer- reira, T.D., Monteiro, C.S., Jorge, P.A., Silva, N.A.: Event-based speckle interrogation for high- bandwidth multi-point optical fiber sensing. arXiv preprint arXiv:2509.22094 (2025)

work page arXiv 2025

[9] [9]

Light: Science and Applications8(2019) https://doi.org/10.1038/ s41377-019-0223-1

Luo, Y., Mengu, D., Yardimci, N.T., Rivenson, Y., Veli, M., Jarrahi, M., Ozcan, A.: Design of task-specific optical systems using broadband diffractive neural networks. Light: Science and Applications8(2019) https://doi.org/10.1038/ s41377-019-0223-1

2019

[10] [10]

Springer (2024)

Fu, T., Zhang, J., Sun, R., Huang, Y., Xu, W., Yang, S., Zhu, Z., Chen, H.: Optical neural net- works: progress and challenges. Springer (2024). https://doi.org/10.1038/s41377-024-01590-3

work page doi:10.1038/s41377-024-01590-3 2024

[11] [11]

Communications Physics8(1), 68 (2025)

Skalli, A., Goldmann, M., Haghighi, N., Reitzen- stein, S., Lott, J.A., Brunner, D.: Annealing- inspired training of an optical neural network with ternary weights. Communications Physics8(1), 68 (2025)

2025

[12] [12]

Computer Communi- cations194, 213–225 (2022) https://doi.org/10

Caiazza, C., Giordano, S., Luconi, V., Vecchio, A.: Edge computing vs centralized cloud: Impact of communication latency on the energy consump- tion of lte terminal nodes. Computer Communi- cations194, 213–225 (2022) https://doi.org/10. 1016/j.comcom.2022.07.026

2022

[13] [13]

IEEE Sensors Journal25, 31545–31562 (2025) https: //doi.org/10.1109/JSEN.2025.3584460

Delaney, E., Brophy, T., Ward, E., Collins, F., Jones, E., Deegan, B., Glavin, M.: Evaluating event-based vision sensing in rain and fog. IEEE Sensors Journal25, 31545–31562 (2025) https: //doi.org/10.1109/JSEN.2025.3584460

work page doi:10.1109/jsen.2025.3584460 2025

[14] [14]

Nature Photonics18, 1076–1082 (2024) https: //doi.org/10.1038/s41566-024-01494-z 7

Yildirim, M., Dinc, N.U., Oguz, I., Psaltis, D., Moser, C.: Nonlinear processing with linear optics. Nature Photonics18, 1076–1082 (2024) https: //doi.org/10.1038/s41566-024-01494-z 7

work page doi:10.1038/s41566-024-01494-z 2024

[15] [15]

eLight4(2024) https://doi.org/10.1186/ s43593-024-00067-5

Hu, J., Liao, K., Din¸ c, N.U., Gigli, C., Bai, B., Gan, T., Li, X., Chen, H., Yang, X., Li, Y., I¸ sıl, Rahman, M.S.S., Li, J., Hu, X., Jarrahi, M., Psaltis, D., Ozcan, A.: Subwavelength imaging using a solid-immersion diffractive optical pro- cessor. eLight4(2024) https://doi.org/10.1186/ s43593-024-00067-5

2024

[16] [16]

Light: Science and Applications7(2018) https://doi

Rahmani, B., Loterie, D., Konstantinou, G., Psaltis, D., Moser, C.: Multimode optical fiber transmission with a deep learning network. Light: Science and Applications7(2018) https://doi. org/10.1038/s41377-018-0074-1

work page doi:10.1038/s41377-018-0074-1 2018

[17] [17]

Optics Express33, 36133 (2025) https://doi.org/ 10.1364/oe.571745

Lan, R., Li, Y., Shen, H., Lu, Y., Hu, H., Zheng, L., Zong, Q., Zang, Y., Zhang, Z.: All-optical color image encryption using multimode fiber speckles and diffractive deep neural networks. Optics Express33, 36133 (2025) https://doi.org/ 10.1364/oe.571745

work page doi:10.1364/oe.571745 2025

[18] [18]

Optics and Laser Technology160(2023) https://doi.org/10.1016/ j.optlastec.2022.109062

Cai, Y., Liu, Y., Li, G., Qin, Q., Pang, L., Ren, W., Wei, J., Wang, M.: Reflective tactile sensor assisted by multimode fiber-based optical cou- pler and fiber specklegram. Optics and Laser Technology160(2023) https://doi.org/10.1016/ j.optlastec.2022.109062

work page arXiv 2023

[19] [19]

Nanomaterials13(2023) https: //doi.org/10.3390/nano13040768

Gu, L., Gao, H., Hu, H.: Demonstration of a learning-empowered fiber specklegram sensor based on focused ion beam milling for refrac- tive index sensing. Nanomaterials13(2023) https: //doi.org/10.3390/nano13040768

work page doi:10.3390/nano13040768 2023

[20] [20]

IEEE Sensors Journal (2025)

Uyar, F., Aldemir, A.S., Kartaloglu, T., Ozbay, E., Ozdur, I.: A compliant mandrel-based fiber optic hydrophone for underwater acoustic sensing. IEEE Sensors Journal (2025)

2025

[21] [21]

IEEE Transactions on Power Delivery25(4), 2526–2534 (2010)

Lima, S.E., Fraz˜ ao, O., Farias, R.G., Ara´ ujo, F.M., Ferreira, L.A., Santos, J.L., Miranda, V.: Mandrel-based fiber-optic sensors for acoustic detection of partial discharges—a proof of con- cept. IEEE Transactions on Power Delivery25(4), 2526–2534 (2010)

2010

[22] [22]

IUCrJ10(5), 568–578 (2023)

Wang, C., Florin, E., Chang, H.-Y., Thayer, J., Yoon, C.H.: Specklenn: a unified embedding for real-time speckle pattern classification in x-ray single-particle imaging with limited labeled exam- ples. IUCrJ10(5), 568–578 (2023)

2023

[23] [23]

Physics Letters A468, 128736 (2023)

Castilho, V., Balthazar, W., Silva, L., Penna, T., Huguenin, J.: Machine learning classification of speckle patterns for roughness measurements. Physics Letters A468, 128736 (2023)

2023

[24] [24]

IEEE Sensors Journal23(18), 20937– 20950 (2023)

Newaz, A., Faruque, M.O., Al Mahmud, R., Sagor, R.H., Khan, M.Z.M.: Machine-learning- enabled multimode fiber specklegram sensors: a review. IEEE Sensors Journal23(18), 20937– 20950 (2023)

2023

[25] [25]

Optics express 23(9), 11898–11911 (2015)

Dr´ emeau, A., Liutkus, A., Martina, D., Katz, O., Sch¨ ulke, C., Krzakala, F., Gigan, S., Daudet, L.: Reference-less measurement of the transmission matrix of a highly scattering material using a dmd and phase retrieval techniques. Optics express 23(9), 11898–11911 (2015)

2015

[26] [26]

SurfaceCodeQuantumCommunication,

Popoff, S.M., Lerosey, G., Carminati, R., Fink, M., Boccara, A.C., Gigan, S.: Measuring the transmission matrix in optics: An approach to the study and control of light propagation in disordered media. Physical Review Letters104 (2010) https://doi.org/10.1103/PhysRevLett.104. 100601 8

work page doi:10.1103/physrevlett.104 2010