Publications de Ali Momeni | EPFL Graph Search

Backpropagation-free Training of Analog AI Accelerators

Romain Christophe Rémy Fleury, Ali Momeni, Matthieu Francis Malléjac

Deep learning has achieved remarkable success in diverse fields in recent years. However, this growth presents significant challenges, particularly in terms of energy consumption during both training and inference phases. While there have been efforts to i ...

Institute of Electrical and Electronics Engineers Inc.2024

Backpropagation-free training of deep physical neural networks

Romain Christophe Rémy Fleury, Babak Rahmani, Ali Momeni, Marc Philipp Del Hougne, Matthieu Francis Malléjac

Recent successes in deep learning for vision and natural language processing are attributed to larger models but come with energy consumption and scalability issues. Current training of digital deep-learning models primarily relies on backpropagation that ...

American Association for the Advancement of Science2023

Parallel temporal signal processing enabled by polarization-multiplexed programmable THz metasurfaces

Romain Christophe Rémy Fleury, Ali Momeni

Under the trends of multifunctionality, tunability, and compactness in modern wave -based signal processors, in this paper, we propose a polarization-multiplexed graphene-based metasurface to realize distinct mathematical operators on the parallel time-dom ...

Optica Publishing Group2022

Parallel wave-based analog computing using metagratings

Romain Christophe Rémy Fleury, Ali Momeni

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 ...

2022

Electromagnetic wave-based extreme deep learning with nonlinear time-Floquet entanglement

Romain Christophe Rémy Fleury, Ali Momeni

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 traditiona ...

Nature Research2022

Switchable and Simultaneous Spatiotemporal Analog Computing with Computational Graphene-based Multilayers

Romain Christophe Rémy Fleury, Ali Momeni

In the past few years, analog computing has experienced rapid development but mostly for a single function. Motivated by parallel space-time computing and miniaturization, we show that reconfigurable graphene-based multilayerss offer a promising path towar ...

2022

Reciprocal Metasurfaces for On-axis Reflective Optical Computing

Karim Achouri, Romain Christophe Rémy Fleury, Ali Momeni

Analog computing has emerged as a promising candidate for real-time and parallel continuous data processing. This paper presents a reciprocal way for realizing asymmetric optical transfer functions (OTFs) in the reflection side of the on-axis processing ch ...

2021

Asymmetric Metal-Dielectric Metacylinders and Their Potential Applications From Engineering Scattering Patterns to Spatial Optical Signal Processing

Romain Christophe Rémy Fleury, Mahdi Safari, Ali Momeni

We propose a bianisotropic hybrid metal-dielectric structure comprising dielectric and metallic cylindrical wedges wherein the composite metacylinder enables advanced control of electric, magnetic, and magnetoelectric resonances. We establish a theoretical ...

American Physical Society (APS)2021

Parallel Analog Computing Based on a 2×2 Multiple-Input Multiple-Output Metasurface Processor With Asymmetric Response

Romain Christophe Rémy Fleury, Ali Momeni, Amirhossein Babaee

We present a polarization-insensitive metasurface processor to perform spatial asymmetric filtering of an incident beam, thereby allowing for real-time parallel analog processing. To enable massive parallel processing, we introduce a multiple-input multipl ...

American Physical Society (APS)2021

Parallel Optical Spatial Signal Processing Based on 2x2 MIMO Computational Metasurface

Romain Christophe Rémy Fleury, Ali Momeni, Amirhossein Babaee

We introduce a novel concept of Multi-Input Multi-Output (MIMO) metasurface processor with asymmetric Optical Transfer Function (OTF) which can perform spatial first-order derivation on two orthogonal distinct input signals for both TM and TE polarizations ...

IEEE2020

Spatial wave control using a self-biased nonlinear metasurface at microwave frequencies

Ali Momeni

Recently, investigation of metasurfaces has been extended to wave control through exploiting nonlinearity. Among all of the ways to achieve tunable metasurfaces with multiplexed performances, nonlinearity is one of the promising choices. Although several p ...

2020

Parallel integro-differential equation solving via multi-channel reciprocal bianisotropic metasurface augmented by normal susceptibilities

Karim Achouri, Ali Momeni

Analog optical signal processing has dramatically transcended the speed and energy limitations accompanied with its digital microelectronic counterparts. Motivated by recent metasurface?s evolution, the angular scattering diversity of a reciprocal passive ...

2019

Generalized Optical Signal Processing Based on Multioperator Metasurfaces Synthesized by Susceptibility Tensors

Karim Achouri, Ali Momeni

This paper theoretically proposes a multichannel nonlocal metasurface computer characterized by generalized sheet transition conditions (GSTCs) and surface susceptibility tensors. The study explores polarization- and angle-multiplexed metasurfaces enabling ...

American Physical Society (APS)2019