arXiv subjects

Romain Fleury

Publications and source records attributed to Romain Fleury.

At least 37 records · Page 2Linked to original sources

arXiv2024

Experimental realization of an active time-modulated acoustic circulator

Reciprocity is one of the fundamental characteristics of wave propagation in linear time-invariant media with preserved time-reversal symmetry. Breaking reciprocity opens the way to numerous applications in the fields of phononics and photonics, as it allows the unidirectional transport of information and energy carried by waves. In acoustics, achieving non-reciprocal behavior remains a challenge, for which time-varying media are one of the solutions. Here, we design and experimentally demonstrate a three-port non-reciprocal acoustic scatterer that behaves as a circulator for audible sound, by actively modulating the effective mass of the acoustic membranes over time. We discuss the conception and experimental validation of such an acoustic circulator, implemented with actively controlled loudspeakers, in the realm of audible and airborne acoustics, and demonstrate its good performance in different scenarios.

physics.app-ph↗

arXiv2024

Training of Physical Neural Networks

Physical neural networks (PNNs) are a class of neural-like networks that leverage the properties of physical systems to perform computation. While PNNs are so far a niche research area with small-scale laboratory demonstrations, they are arguably one of the most underappreciated important opportunities in modern AI. Could we train AI models 1000x larger than current ones? Could we do this and also have them perform inference locally and privately on edge devices, such as smartphones or sensors? Research over the past few years has shown that the answer to all these questions is likely "yes, with enough research": PNNs could one day radically change what is possible and practical for AI systems. To do this will however require rethinking both how AI models work, and how they are trained - primarily by considering the problems through the constraints of the underlying hardware physics. To train PNNs at large scale, many methods including backpropagation-based and backpropagation-free approaches are now being explored. These methods have various trade-offs, and so far no method has been shown to scale to the same scale and performance as the backpropagation algorithm widely used in deep learning today. However, this is rapidly changing, and a diverse ecosystem of training techniques provides clues for how PNNs may one day be utilized to create both more efficient realizations of current-scale AI models, and to enable unprecedented-scale models.

physics.app-ph↗

arXiv2024

Renormalization group of topological scattering networks

Exploring and understanding topological phases in systems with strong distributed disorder requires developing fundamentally new approaches to replace traditional tools such as topological band theory. Here, we present a general real-space renormalization group (RG) approach for scattering models, which is capable of dealing with strong distributed disorder without relying on the renormalization of Hamiltonians or wave functions. Such scheme, based on a block-scattering transformation combined with a replica strategy, is applied for a comprehensive study of strongly disordered unitary scattering networks with localized bulk states, uncovering a connection between topological physics and critical behavior. Our RG scheme leads to topological flow diagrams that unveil how the microscopic competition between reflection and non-reciprocity leads to the large-scale emergence of macroscopic scattering attractors, corresponding to trivial and topological insulators. Our findings are confirmed by a scaling analysis of the localization length (LL) and critical exponents, and experimentally validated. The results not only shed light on the fundamental understanding of topological phase transitions and scaling properties in strongly disordered regimes, but also pave the way for practical applications in modern topological condensed-matter and photonics, where disorder may be seen as a useful design degree of freedom, and no longer as a hindrance.

cond-mat.dis-nn↗

arXiv2024

Synchronization-based lossless non-reciprocal scattering

Breaking the reciprocity of wave propagation is a problem of fundamental interest, and a mucht-sought functionality in practical applications, both in photonics and phononics. Although it has been achieved using resonant linear scattering from cavities with broken time-reversal symmetry, such realizations have remained inescapably plagued by inherent passivity constraints, which make absorption losses unavoidable, leading to stringent limitations in transmitted power. In this work, we solve this problem by converting the cavity resonance into a limit cycle, exploiting the uncharted interplay between non-linearity, gain, and non-reciprocity. Remarkably, strong enough incident waves can synchronize with these self-sustained oscillations and use their energy for amplification. We theoretically and experimentally demonstrate that this mechanism can simultaneously enhance non-reciprocity and compensate absorption. Real-world acoustic scattering experiments allow us to observe perfect non-reciprocal transmission of audible sound in a synchronisation-based 3-port circulator with full immunity against losses.

physics.app-ph↗

arXiv2024

Experimental observation of topological transition in linear and non-linear parametric oscillators

Parametric oscillators are examples of externally driven systems that can exhibit two stable states with opposite phase depending on the initial conditions. In this work, we propose to study what happens when the external forcing is perturbed by a continuously parametrized defect. Initially in one of its stable state, the oscillator will be perturbed by the defect and finally reach another stable state, which can be its initial one or the other one. For some critical value of the defect parameter, the final state changes abruptly. We investigate theoretically and experimentally such transition both in the linear and non-linear case, and the effect of non-linearities is discussed. A topological interpretation in terms of winding number is proposed, and we show that winding changes correspond to singularities in the temporal dynamics. An experimental observation of such transition is performed using parametric Faraday instability at the surface of a vibrated fluid.

nlin.PS↗

arXiv2024

Experimental observation of parabolic wakes in thin plates

Wakes are medium perturbations created by a moving object, such as wave patterns behind boats, or wingtip vortices following an aircraft. Here, we report about an experimental study of an uncharted form of parabolic wakes occurring in media with the group velocity twice larger than the phase velocity, as opposed to the conventional case of Kelvin wakes. They are formed by moving a laser spot on a thin plate, which excites a unique wake pattern made of confocal parabolas, due to the quadratic dispersion of the zero-order flexural Lamb mode. If the spatial dimensions are rescaled by the perturbation velocity and material constant, we obtain a single universal wake with constant parabolic focal lengths. We demonstrate an evanescent regime above the critical frequency where the wave components oscillate exclusively in the direction parallel to the perturbation path, with an opening angle of 90°. We define a dimensionless number analogous to Froude and Mach numbers, which determines whether the complete parabolic wake pattern will be excited by the moving source or not.

physics.app-ph↗

arXiv2023

Wave momentum shaping for moving objects in heterogeneous and dynamic media

Light and sound waves have the fascinating property that they can move objects through the transfer of linear or angular momentum. This ability has led to the development of optical and acoustic tweezers, with applications ranging from biomedical engineering to quantum optics. Although impressive manipulation results have been achieved, the stringent requirement for a highly controlled, low-reverberant, and static environment still hinders the applicability of these techniques in many scenarios. Here, we overcome this challenge and demonstrate the manipulation of objects in disordered and dynamic media, by optimally tailoring the momentum of sound waves iteratively in the far field. The method does not require information about the object's physical properties or the spatial structure of the surrounding medium but relies only on a real-time scattering matrix measurement and a positional guidestar. Our experiment demonstrates the possibility of optimally moving and rotating objects, extending the reach of wave-based object manipulation to complex and dynamic scattering media. We envision new opportunities for biomedical applications, sensing, or manufacturing.

physics.app-ph↗

arXiv2023

Backpropagation-free Training of Deep Physical Neural Networks

Recent years have witnessed the outstanding success of deep learning in various fields such as vision and natural language processing. This success is largely indebted to the massive size of deep learning models that is expected to increase unceasingly. This growth of the deep learning models is accompanied by issues related to their considerable energy consumption, both during the training and inference phases, as well as their scalability. Although a number of work based on unconventional physical systems have been proposed which addresses the issue of energy efficiency in the inference phase, efficient training of deep learning models has remained unaddressed. So far, training of digital deep learning models mainly relies on backpropagation, which is not suitable for physical implementation as it requires perfect knowledge of the computation performed in the so-called forward pass of the neural network. Here, we tackle this issue by proposing a simple deep neural network architecture augmented by a biologically plausible learning algorithm, referred to as "model-free forward-forward training". The proposed architecture enables training deep physical neural networks consisting of layers of physical nonlinear systems, without requiring detailed knowledge of the nonlinear physical layers' properties. We show that our method outperforms state-of-the-art hardware-aware training methods by improving training speed, decreasing digital computations, and reducing power consumption in physical systems. We demonstrate the adaptability of the proposed method, even in systems exposed to dynamic or unpredictable external perturbations. To showcase the universality of our approach, we train diverse wave-based physical neural networks that vary in the underlying wave phenomenon and the type of non-linearity they use, to perform vowel and image classification tasks experimentally.

cs.LG↗

arXiv2023

Active Su-Schrieffer-Heeger-like System

An acoustic dimer composed of two electronically controlled electro-acoustic resonators is presented in view of exploring one-dimensional topological phenomena.

physics.app-ph↗

arXiv2023

Scattering from Time-modulated Transmission Line Loads: Theory and Experiments in Acoustics

Scattering wave systems that are periodically modulated in time offer many new degrees of freedom to control waves both in spatial and frequency domains. Such systems, albeit linear, do not conserve frequency and require the adaptation of the usual theories and methods. In this paper, we provide a general extension of transmission line or telegraph equations to periodically time-modulated systems. As a by-product of the theory, we obtain a general approach to compute and measure the complete scattering matrix of such systems. Finally, the proposed theory and methods are applied and validated on a concrete practical example in the realm of airborne acoustics: a time-modulated actively controlled loudspeaker membrane terminating a monomode waveguide. Different modulation functions and parameters are tested. The experimental results are compared to both numerical simulation and an analytical model based on a two time-scale method.

physics.app-ph↗

arXiv2023

Physics-inspired Neuroacoustic Computing Based on Tunable Nonlinear Multiple-scattering

Waves, such as light and sound, inherently bounce and mix due to multiple scattering induced by the complex material objects that surround us. This scattering process severely scrambles the information carried by waves, challenging conventional communication systems, sensing paradigms, and wave-based computing schemes. Here, we show that instead of being a hindrance, multiple scattering can be beneficial to enable and enhance analog nonlinear information mapping, allowing for the direct physical implementation of computational paradigms such as reservoir computing and extreme learning machines. We propose a physics-inspired version of such computational architectures for speech and vowel recognition that operate directly in the native domain of the input signal, namely on real-sounds, without any digital pre-processing or encoding conversion and backpropagation training computation. We first implement it in a proof-of-concept prototype, a nonlinear chaotic acoustic cavity containing multiple tunable and power-efficient nonlinear meta-scatterers. We prove the efficiency of the acoustic-based computing system for vowel recognition tasks with high testing classification accuracy (91.4%). Finally, we demonstrate the high performance of vowel recognition in the natural environment of a reverberation room. Our results open the way for efficient acoustic learning machines that operate directly on the input sound, and leverage physics to enable Natural Language Processing (NLP).

cs.SD↗

arXiv2023

Anomalous Floquet Topological Disclination States

Recently, non-reciprocal two-dimensional unitary scattering networks have gained considerable interest due to the possibility of obtaining robust edge wave propagation in the anomalous Floquet phase. Conversely, zero-dimensional topological states in such networks have been left uncharted. Here, we demonstrate the existence of Floquet disclination states in non-reciprocal scattering networks. The disclination states, characterized by spectral charges, nucleate in the anomalous phase from a resonant rotation-symmetric phase matching condition, and survive until the bandgaps close. Once coupled to the radiation continuum feeding the anomalous chiral edge state, they can induce intriguing topological disclination bound states in the continuum (BICs), associated with an extreme confinement and lifetime. Altogether, anomalous Floquet disclination states and topological disclination BIC broaden the applications of disclination states to microwave, acoustic or optical scattering networks, with new possibilities in chiral topological lasing, robust energy squeezing from topological bound states, and switchable lasing and anti-lasing behavior induced via unidirectional topological coupling.

cond-mat.mes-hall↗

arXiv2023

Extreme Spatial Dispersion in Nonlocally-Resonant Elastic Metamaterials

To date, the vast majority of architected materials have leveraged two physical principles to control wave behavior, namely Bragg interference and local resonances. Here, we describe a third path: structures that accommodate a finite number of delocalized zero-energy modes, leading to anomalous dispersion cones that nucleate from extreme spatial dispersion at 0 Hz. We explain how to design such zero-energy modes in the context of elasticity and show that many of the landmark wave properties of metamaterials can also be induced at an extremely subwavelength scale by the associated anomalous cones, without suffering from the same bandwidth limitations. We then validate our theory through a combination of simulations and experiments. Finally, we present an inverse design method to produce anomalous cones at desired locations in momentum space.

physics.app-ph↗

arXiv2022

Anomalous topological waves in strongly amorphous scattering networks

Topological insulators are crystalline materials that have revolutionized our ability to control wave transport. They provide us with unidirectional channels that are immune to obstacles, defects or local disorder, and can even survive some random deformations of their crystalline structures. However, they always break down when the level of disorder or amorphism gets too large, transitioning to a topologically trivial Anderson insulating phase. Here, we demonstrate a two-dimensional amorphous topological regime that survives arbitrarily strong levels of amorphism. We implement it for electromagnetic waves in a non-reciprocal scattering network and experimentally demonstrate the existence of unidirectional edge transport in the strong amorphous limit. This edge transport is shown to be mediated by an anomalous edge state whose topological origin is evidenced by direct topological invariant measurements. Our findings extend the reach of topological physics to a new class of systems in which strong amorphism can induce, enhance and guarantee the topological edge transport instead of impeding it.

physics.app-ph↗

arXiv2022

Ultrabroadband sound control with deep-subwavelength plasmacoustic metalayers

Controlling audible sound requires inherently broadband and subwavelength acoustic solutions, which are to date, crucially missing. This includes current noise absorption methods, such as porous materials or acoustic resonators, which are typically inefficient below 1 kHz, or fundamentally narrowband. Here, we solve this vexing issue by introducing the concept of plasmacoustic metalayers. We demonstrate that the dynamics of small layers of air plasma can be controlled to interact with sound in an ultrabroadband way and over deep-subwavelength distances. Exploiting the unique physics of plasmacoustic metalayers, we experimentally demonstrate perfect sound absorption and tunable acoustic reflection over more than two frequency decades, from several Hz to the kHz range, with transparent plasma layers of thicknesses down to $λ/1000$. Such unprecedented bandwidth and compactness opens new doors in a variety of applications, including noise control, audio-engineering, room acoustics, imaging and metamaterial design.

physics.app-ph↗

arXiv2021

A single metagrating metastructure for wave-based parallel analog computing

Wave-based signal processing has witnessed a significant expansion of interest in a variety of science and engineering disciplines, as it provides new opportunities for achieving high-speed and low-power operations. Although flat optics desires integrable components to perform multiple missions, yet, the current wave-based analog computers can engineer only the spatial content of the input signal where the processed signal obeys the traditional version of Snell's law. In this paper, we propose a multi-functional metagrating to modulate both spatial and angular properties of the input signal whereby both symmetric and asymmetric optical transfer functions are realized using high-order space harmonics. The performance of the designed compound metallic grating is validated through several investigations where closed-form expressions are suggested to extract the phase and amplitude information of the diffractive modes. Several illustrative examples are demonstrated to show that the proposed metagrating allows for simultaneous parallel analog computing tasks such as first- and second-order spatial differentiation through a single multi-channel structured surface. It is anticipated that the designed platform brings a new twist to the field of optical signal processing and opens up large perspectives for simple integrated image processing systems.

physics.optics↗

arXiv2021

Wave-based extreme deep learning based on non-linear time-Floquet entanglement

Wave-based analog signal processing holds the promise of extremely fast, on-the-fly, power-efficient data processing, occurring as a wave propagates through an artificially engineered medium. Yet, due to the fundamentally weak non-linearities of traditional wave materials, such analog processors have been so far largely confined to simple linear projections such as image edge detection or matrix multiplications. Complex neuromorphic computing tasks, which inherently require strong non-linearities, have so far remained out-of-reach of wave-based solutions, with a few attempts that implemented non-linearities on the digital front, or used weak and inflexible non-linear sensors, restraining the learning performance. Here, we tackle this issue by demonstrating the relevance of Time-Floquet physics to induce a strong non-linear entanglement between signal inputs at different frequencies, enabling a power-efficient and versatile wave platform for analog extreme deep learning involving a single, uniformly modulated dielectric layer and a scattering medium. We prove the efficiency of the method for extreme learning machines and reservoir computing to solve a range of challenging learning tasks, from forecasting chaotic time series to the simultaneous classification of distinct datasets. Our results open the way for wave-based machine learning with high energy efficiency, speed, and scalability.

cs.ET↗

arXiv2021

Time-varying components for enhancing wireless transfer of power and information

Temporal modulation of components of electromagnetic systems provides an exceptional opportunity to engineer the response of those systems in a desired fashion, both in the time and frequency domains. For engineering time-modulated systems, one needs to thoroughly study the basic concepts and understand the salient characteristics of temporal modulation. In this paper, we carefully study physical models of basic bulk circuit elements -- capacitors, inductors, and resistors -- as frequency dispersive and time-varying components and study their effects in the case of periodical time modulations. We develop a solid theory for understanding these elements, and apply it to two important applications: wireless power transfer and antennas. For the first application, we show that by periodically modulating the mutual inductance between the transmitter and receiver, the fundamental limits of classical wireless power transfer systems can be overcome. Regarding the second application, we consider a time-varying source for electrically small dipole antennas and show how time modulation can enhance the antenna performance. The developed theory of electromagnetic systems engineered by temporal modulation is applicable from radio frequencies to optical wavelengths.

physics.app-ph↗