arXiv subjects

Sylvain Gigan

Publications and source records attributed to Sylvain Gigan.

At least 19 recordsLinked to original sources

arXiv2026

Quantifying structural nonlinearity in a disordered Fabry-Pérot cavity

Structural nonlinearity is the emergence of nonlinear input-output transformations from purely linear processes. In optics, this arises from the multiple interactions of light with input data. Current implementations achieve either a finite number of interactions, or partial modulation of the optical field. In this study, we present an architecture that combines both arbitrarily many interactions and full modulation. Our design consists of a Fabry-Pérot cavity, where one mirror is replaced by a spatial light modulator on which data is displayed. We develop an analytical model from which we derive closed-form expressions for the main metrics of nonlinearity, and find good agreement with experiment. Our analysis extends beyond the present implementation and reveals relationships that are independent of a specific physical system, providing a basis for a unified description of structural nonlinearity, and a first-principles design guide for machine learning applications.

physics.optics↗

arXiv2026

Reconfigurable Optical Platform for One-way Quantum Communication Complexity

Demonstrating a practical quantum advantage remains a central goal in quantum information science. While quantum computational supremacy is still technologically demanding, communication complexity offers a promising route to showcase quantum advantage with current photonic platforms. Here we introduce a reconfigurable optical platform for one-way quantum communication complexity based on multimode fibers and wavefront shaping. We experimentally validate it by implementing a genuine one-way quantum communication complexity problem for which an exponential quantum--classical communication separation is known. Complementary numerical simulations show that the same reconfigurable decoding architecture can support more general one-way communication tasks with comparable performance, while also offering a route to higher-dimensional implementations without increasing hardware complexity. Together, these results establish multimode-fiber wavefront shaping as a versatile hardware platform for one-way quantum communication complexity and provide a concrete roadmap toward more demanding protocols, where stronger quantum--classical separations could enable practical demonstrations of quantum advantage.

quant-ph↗

arXiv2026

Low Photon Number Non-Invasive Imaging Through Time-Varying Diffusers

Optical imaging plays a crucial role in advancing science and technology, enabling applications in fields ranging from biomedicine to astronomy. However, imaging through scattering media such as biological tissues, fog, or turbulent atmosphere remains a major challenge. Light scattering and absorption in such media make imaging challenging; in the case of time-varying scatterers and low light regime imaging of incoherent objects has not been demonstrated so far. We present the first demonstration of such non-invasive imaging of dim objects hidden behind dynamic scattering layers, obtaining robust reconstruction even at extremely low photon counts per frame. We achieve this by developing a new data-processing approach. In our experiment, we utilize a photon number resolving camera to capture a sequence of frames, containing on average, fewer than one photon per pixel. We validate our approach in microscopy, where we reconstruct images of biological samples stained with standard fluorescent dyes. Beyond microscopy, our approach can be applied in different imaging techniques, such as endoscopy based on multicore fibers or ground-based astronomical observations.

physics.optics↗

arXiv2026

TRON: Trainable, architecture-reconfigurable random optical neural networks

Deep learning has triggered explosive growth in the demand for specialized hardware processors, thus motivating the development of scalable and reconfigurable computing substrates. Optical processors offer a fundamentally different computing paradigm, combining massive parallelism and ultrahigh bandwidth with the potential for substantial energy savings. However, progress has been constrained by the absence of scalable and reconfigurable architectures that can implement a broad class of network architectures. Here, we introduce TRON, a scalable and trainable optoelectronic deep optical neural network that exploits a multi-scattering medium and a DMD as a learnable, high-dimensional dense optical matrix multiplier, processing with fixed and tunable optical operations. We perform in-situ optimization of the optical parameters involved in the scattering process, together with automated neural architecture search (NAS) and optimization directly on optics. The experimental results demonstrate that in-situ NAS is essential to discover architectures that adapt to both the task and hardware constraints, establishing a viable path towards large-scale optical processors for next-generation machine learning and data-intensive computing.

physics.optics↗

arXiv2026

High-resolution scanning fluorescence imaging through scattering via speckle replica alignment and variance computation

Fluorescence imaging is an essential diagnostic tool in many fields, but diffraction-limited optical imaging at depth is limited by scattering. Here, we present a method based on multiple random illuminations, combined with a computational framework that retrieves high-resolution images by aligning local speckle replicas and computing their pixel-wise variance. We demonstrate its versatility in two regimes: linear wide-field one-photon (1P) fluorescence imaging and nonlinear two-photon (2P) fluorescence imaging where the object is excited by a scanned speckle field and detected with a single-pixel detector. This approach outperforms standard autocorrelation techniques in terms of resolution and convergence.

physics.optics↗

arXiv2026

Real-time Calibration-free Imaging Through Dynamic and Distinct Multimode Fibers via Spatial Harmonic Invariant Nonlinear Encoding (SHINE)

Multimode fibers (MMFs) provide a compact, high-throughput platform for minimally invasive imaging and information transmission. However, their utility is fundamentally constrained by mode mixing, which renders image transmission spatially disrupted and sensitive to external perturbations. Current imaging methods typically rely on transmission matrix measurement or deep learning models that are fragile to fiber movement, necessitating frequent, time-consuming calibrations and re-calibrations that are easily disrupted and fail to generalize across different fiber configurations, let alone across entirely distinct fibers. Here, we propose a calibration and feedback-free MMF coherent imaging paradigm, that we termed Spatial Harmonic Invariant Nonlinear Encoding (SHINE). By leveraging the angle-dependent phase-matching conditions of second-harmonic generation, we encode spatial features into broadband spectral signatures that possess intrinsic insensitivity not only to modal scrambling but also to fiber bending, movement, as well as structural variations. This spectral representation enables a deep learning model to robustly reconstruct images in real time despite dynamic perturbations and even generalizes well to distinct MMFs without recalibration or feedback. We achieve experimentally an average Pearson correlation coefficient (PCC) of 0.82 for image reconstruction tasks on Fashion-MNIST and a classification accuracy of 92.3% on HERLEV biomedical dataset. Uniquely, our method exhibits remarkable cross-fiber generalization: a model trained on a single MMF successfully reconstructs images transmitted through entirely distinct, previously unseen MMFs with a PCC of 0.74. These results establish a robust, calibration-free framework for imaging through MMFs in real time, paving the way for practical, resilient optical diagnostics that operate without distal-end feedback.

physics.optics↗

arXiv2026

Training deep physical neural networks with local physical information bottleneck

Deep learning has revolutionized modern society but faces growing energy and latency constraints. Deep physical neural networks (PNNs) are interconnected computing systems that directly exploit analog dynamics for energy-efficient, ultrafast AI execution. Realizing this potential, however, requires universal training methods tailored to physical intricacies. Here, we present the Physical Information Bottleneck (PIB), a general and efficient framework that integrates information theory and local learning, enabling deep PNNs to learn under arbitrary physical dynamics. By allocating matrix-based information bottlenecks to each unit, we demonstrate supervised, unsupervised, and reinforcement learning across electronic memristive chips and optical computing platforms. PIB also adapts to severe hardware faults and allows for parallel training via geographically distributed resources. Bypassing auxiliary digital models and contrastive measurements, PIB recasts PNN training as an intrinsic, scalable information-theoretic process compatible with diverse physical substrates.

cs.LG↗

arXiv2026

Light and Sound Driven Wavefront Shaping and Imaging through Scattering Tissue

Deep, high-resolution imaging is essential for unraveling biological complexity and advancing medical diagnostics, yet scattering fundamentally limits optical methods. Among the most promising approaches, photoacoustic imaging achieves penetration into deep tissue but with coarse resolution, while fluorescence provides subcellular detail but is confined to shallow depths. This depth-resolution trade-off remains a central barrier to biomedical imaging. To bridge this fundamental gap, we present a hybrid dual-modal strategy that combines the benefits of photoacoustic and fluorescence modalities. Our approach leverages hybrid opto-acoustic feedback for wavefront shaping and computational imaging through scattering media. By combining these complementary signals into a nonlinear feedback metric, we achieve robust optical focusing even under signal degradation. In particular, we show that photoacoustic-guided wavefront shaping inherently generates fluorescence that can be harvested for computational high-resolution imaging even within highly scattering biological tissues, thereby leveraging the complementary strengths of both modalities in a single framework. Proof-of-concept experiments demonstrate this synergistic approach, paving the way for optical imaging techniques that fully leverage the potential of such dual-modalities for large depth penetration and high resolution in complex biological tissues.

physics.optics↗

arXiv2026

Controlling microalgae populations by phototactic memory

Understanding how microorganisms navigate in complex environments is a central question in active matter and biological physics. Phototaxis - the ability to use light as a navigation cue - is a widespread strategy in motile microalgae to optimise photosynthesis and avoid light-induced stress. The microalga Chlamydomonas reinhardtii is a model system for studying this behaviour, where navigation is classically attributed to a photosensitive organelle named eyespot. While this mechanism enables cells to sense the direction of incoming light, their response to light intensity gradients remains less understood. Here we show that structured light landscapes can guide microalgae populations and localise them in defined spatial regions. By analysing single-cell trajectories, we find that cells actively steer relative to the local light gradient, and a comparison with a minimal theoretical model shows that a short-time memory of light exposure acting on the transition between positive and negative phototaxis is necessary to reproduce the observed accumulation. At longer times, we observe a gradual decrease in cell number density within the trapping region, consistent with phototactic adaptation. Beyond controlling population dynamics, our results reveal new aspects of phototactic behaviour, highlighting gradient-aligned steering together with temporal integration as central mechanisms for navigation in structured environments.

cond-mat.soft↗

arXiv2025

Terahertz Fourier Ptychographic Imaging

High-resolution imaging in the terahertz (THz) spectral range remains fundamentally constrained by the limited numerical apertures of currently existing state-of-the-art imagers, which restricts its applicability across many fields, such as imaging in complex media or nondestructive testing. To address this challenge, we introduce a proof-of-concept implementation of THz Fourier Ptychographic imaging to enhance spatial resolution without requiring extensive hardware modifications. Our method employs a motorized kinematic mirror to generate a sequence of controlled, multi-angle plane-wave illuminations, with each resulting oblique-illumination intensity image encoding a limited portion of the spatial-frequency content of the target imaging sample. These measurements are combined in the Fourier domain using an aberration-corrected iterative phase-retrieval algorithm integrated with an efficient illumination calibration scheme, which enables the reconstruction of resolution-enhanced amplitude and phase images through the synthetic expansion of the effective numerical aperture. Our work establishes a robust framework for high-resolution THz imaging and paves the way for a wide array of applications in materials characterization, spectroscopy, and non-destructive evaluation.

physics.optics↗

arXiv2025

Optical kernel machine with programmable nonlinearity

Optical kernel machines offer high throughput and low latency. A nonlinear optical kernel can handle complex nonlinear data, but power consumption is typically high with the conventional nonlinear optical approach. To overcome this issue, we present an optical kernel with structural nonlinearity that can be continuously tuned at low power. It is implemented in a linear optical scattering cavity with a reconfigurable micro-mirror array. By tuning the degree of nonlinearity with multiple scattering, we vary the kernel sensitivity and information capacity. We further optimize the kernel nonlinearity to best approximate the parity functions from first order to fifth order for binary inputs. Our scheme offers potential applicability across photonic platforms, providing programmable kernels with high performance and low power consumption.

physics.optics↗

arXiv2025

Optical Computing with Spectrally Multiplexed Features in Complex Media

Artificial intelligence (AI) has rapidly evolved into a critical technology; however, electrical hardware struggles to keep pace with the exponential growth of AI models. Free space optical hardware provides alternative approaches for large-scale optical processing, and in-memory computing, with applications across diverse machine learning tasks. Here, we explore the use of broadband light scattering in free-space optical components, specifically complex media, which generate uncorrelated optical features at each wavelength. By treating individual wavelengths as independent predictors, we demonstrate improved classification accuracy through in-silico majority voting, along with the ability to estimate uncertainty without requiring access to the model's probability outputs. We further demonstrate that linearly combining multiwavelength features, akin to spectral shaping, enables us to tune output features with improved performance on classification tasks, potentially eliminating the need for multiple digital post-processing steps. These findings illustrate the spectral multiplexing or broadband advantage for free-space optical computing.

physics.optics↗

arXiv2025

Harnessing optical disorder for Bell inequalities violation

Bell inequalities are a cornerstone of quantum physics. By carefully selecting measurement bases (typically polarization), their violation certifies quantum entanglement. Such measurements are disrupted by the presence of optical disorder in propagation paths, including polarization or spatial mode mixing in fibers and through free-space turbulence. Here, we demonstrate that disorder can instead be exploited as a resource to certify entanglement via a Bell inequality test. In our experiment, one photon of a polarization-entangled pair propagates through a commercial multimode fiber that scrambles spatial and polarization modes, producing a speckle pattern, while the other photon remains with the sender. By spatially resolving the speckle intensity pattern, we naturally access a large set of random and unknown polarization projections. We show that this set is statistically sufficient to violate a Bell inequality, thereby certifying entanglement without requiring active correction techniques. Our approach provides a fundamentally new way to test Bell inequalities, eliminating the need for an explicit choice of measurement basis, and offering a practical solution for entanglement certification in real-world quantum communication channels where disorder is unavoidable.

quant-ph↗

arXiv2025

Non-classical optimization of entangled photons through complex media

Optimization approaches are ubiquitous in physics. In optics, they are key to manipulating light through complex media, enabling applications ranging from imaging to photonic simulators. In most demonstrations, however, the optimization process is implemented using classical coherent light, leading to a purely classical solution. Here we introduce the concept of optical non-classical optimization in complex media. We experimentally demonstrate the control and refocusing of non-classical light -- namely, entangled photon pairs -- through a scattering medium by directly optimizing the output coincidence rate. The optimal solutions found with this approach differ from those obtained using classical optimization, a result of entanglement in the input state. Beyond imaging, this genuinely non-classical optimization method has potential to tackle complex problems, as we show by simulating a spin-glass model with multi-spin interactions.

quant-ph↗

arXiv2025

Two-photon microscopy through scattering media harnessing speckle autocorrelation

Two-photon (2P) microscopy is a powerful technique for deep-tissue fluorescence imaging; however, tissue scattering limits its effectiveness for depth imaging using conventional approaches. Despite typical strategies having been put forward to extend depth imaging capabilities based on wave-front shaping (WFS), computationally recovering images remains a significant challenge using 2P signal. In this work, we demonstrate the successful reconstruction of fluorescent objects behind scattering layers using 2P microscopy, utilizing the optical memory effect (ME) along with the speckle autocorrelation technique and a phase retrieval algorithm. Our results highlight the effectiveness of this method, offering significant potential for improving depth imaging capabilities in 2P microscopy through scattering media.

physics.optics↗

arXiv2025

Harnessing Photon Indistinguishability in Quantum Extreme Learning Machines

Recent advancements in machine learning have led to an exponential increase in computational demands, driving the need for innovative computing platforms. Quantum computing, with its Hilbert space scaling exponentially with the number of particles, emerges as a promising solution. In this work, we implement a quantum extreme machine learning (QELM) protocol leveraging indistinguishable photon pairs and multimode fiber as a random densly connected layer. We experimentally study QELM performance based on photon coincidences -- for distinguishable and indistinguishable photons -- on an image classification task. Simulations further show that increasing the number of photons reveals a clear quantum advantage. We relate this improved performance to the enhanced dimensionality and expressivity of the feature space, as indicated by the increased rank of the feature matrix in both experiment and simulation.

quant-ph↗

arXiv2025

Streamlined optical training of large-scale modern deep learning architectures with direct feedback alignment

Modern deep learning relies nearly exclusively on dedicated electronic hardware accelerators. Photonic approaches, with low consumption and high operation speed, are increasingly considered for inference but, to date, remain mostly limited to relatively basic tasks. Simultaneously, the problem of training deep and complex neural networks, overwhelmingly performed through backpropagation, remains a significant limitation to the size and, consequently, the performance of current architectures and a major compute and energy bottleneck. Here, we experimentally implement a versatile and scalable training algorithm, called direct feedback alignment, on a hybrid electronic-photonic platform. An optical processing unit performs large-scale random matrix multiplications, which is the central operation of this algorithm, at speeds up to 1500 TeraOPS under 30 Watts of power. We perform optical training of modern deep learning architectures, including Transformers, with more than 1B parameters, and obtain good performances on language, vision, and diffusion-based generative tasks. We study the scaling of the training time, and demonstrate a potential advantage of our hybrid opto-electronic approach for ultra-deep and wide neural networks, thus opening a promising route to sustain the exponential growth of modern artificial intelligence beyond traditional von Neumann approaches.

cs.ET↗

arXiv2025

Three-dimensional holographic imaging of incoherent objects through scattering media

Three-dimensional (3D) high-resolution imaging is essential in microscopy, yet light scattering poses significant challenges in achieving it. Here, we present an approach to holographic imaging of spatially incoherent objects through scattering media, utilizing a virtual medium that replicates the scattering effects of the actual medium. This medium is constructed by retrieving mutually incoherent fields from the object, and exploiting the spatial correlations between them. By numerically propagating the incoherent fields through the virtual medium, we non-invasively compensate for scattering, achieving accurate 3D reconstructions of hidden objects. Experimental validation with fluorescent and synthetic incoherent objects confirms the effectiveness of this approach, opening new possibilities for advanced 3D high-resolution microscopy in scattering environments.

physics.optics↗