Babak Rahmani
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Babak Rahmani has a total of 72 co-authors. Here are the top co-authors by number of shared publications:
Christophe Moser
École Polytechnique Fédérale de Lausanne, Laboratory of Applied Photonics Devices, CH-1015 Lausanne, Switzerland
5 shared publicationsDamien Loterie
Readily3D, CH-1015 Lausanne
5 shared publicationsDemetri Psaltis
Ecole Polytechnique Fédérale de Lausanne, Laboratory of Optics, CH-1015 Lausanne, Switzerland
5 shared publications- Christophe Moser
Laboratory of Applied Photonics Devices, School of Engineering, École Polytechnique Fédérale de Lausanne, CH-1015, Lausanne, Switzerland
4 shared publications - Demetri Psaltis
Optics Laboratory (LO), École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland
4 shared publications Eirini Kakkava
Ecole Polytechnique Fédérale de Lausanne, Laboratory of Optics, CH-1015 Lausanne, Switzerland
4 shared publications
Latest Publications
- Top Cited PapersOpen Access
- 3 September 2025
- Published by Springer Nature in Nature
- Vol. 645 (8080) , 1-8
Analog optical computer for AI inference and combinatorial optimization
Artificial intelligence (AI) and combinatorial optimization drive applications across science and industry, but their increasing energy demands challenge the sustainability of digital computing. Most unconventional computing systems1–7 target either AI or optimization workloads and rely on frequent, energy-intensive digital conversions, limiting efficiency. These systems also face application-hardware mismatches, whether handling memory-bottlenecked neural models, mapping real-world optimization problems or contending with inherent analog noise. Here we introduce an analog optical computer (AOC) that combines analog electronics and three-dimensional optics to accelerate AI inference and combinatorial optimization in a single platform. This dual-domain capability is enabled by a rapid fixed-point search, which avoids digital conversions and enhances noise robustness. With this fixed-point abstraction, the AOC implements emerging compute-bound neural models with recursive reasoning potential and realizes an advanced gradient-descent approach for expressive optimization. We demonstrate the benefits of co-designing the hardware and abstraction, echoing the co-evolution of digital accelerators and deep learning models, through four case studies: image classification, nonlinear regression, medical image reconstruction and financial transaction settlement. Built with scalable, consumer-grade technologies, the AOC paves a promising path for faster and sustainable computing. Its native support for iterative, compute-intensive models offers a scalable analog platform for fostering future innovation in AI and optimization.
- 3 September 2025
- Published by Springer Nature in Nature
- Vol. 645 (8080) , 1-8
Top Cited PapersOpen AccessAnalog optical computer for AI inference and combinatorial optimization
Artificial intelligence (AI) and combinatorial optimization drive applications across science and industry, but their increasing energy demands challenge the sustainability of digital computing. Most unconventional computing systems1–7 target either AI or optimization workloads and rely on frequent, energy-intensive digital conversions, limiting efficiency. These systems also face application-hardware mismatches, whether handling memory-bottlenecked neural models, mapping real-world optimization problems or contending with inherent analog noise. Here we introduce an analog optical computer (AOC) that combines analog electronics and three-dimensional optics to accelerate AI inference and combinatorial optimization in a single platform. This dual-domain capability is enabled by a rapid fixed-point search, which avoids digital conversions and enhances noise robustness. With this fixed-point abstraction, the AOC implements emerging compute-bound neural models with recursive reasoning potential and realizes an advanced gradient-descent approach for expressive optimization. We demonstrate the benefits of co-designing the hardware and abstraction, echoing the co-evolution of digital accelerators and deep learning models, through four case studies: image classification, nonlinear regression, medical image reconstruction and financial transaction settlement. Built with scalable, consumer-grade technologies, the AOC paves a promising path for faster and sustainable computing. Its native support for iterative, compute-intensive models offers a scalable analog platform for fostering future innovation in AI and optimization.
- Top Cited Papers
- 3 September 2025
- Published by Springer Nature in Nature
- Vol. 645 (8079) , 53-61
Training of physical neural networks
Physical neural networks (PNNs) are a class of neural-like networks that make use of analogue physical systems to perform computations. Although at present confined to small-scale laboratory demonstrations, PNNs could one day transform how artificial intelligence (AI) calculations are performed. Could we train AI models many orders of magnitude larger than present ones? Could we perform model inference locally and privately on edge devices? Research over the past few years has shown that the answer to these questions is probably “yes, with enough research”. Because PNNs can make use of analogue physical computations more directly, flexibly and opportunistically than traditional computing hardware, they could change what is possible and practical for AI systems. To do this, however, will require notable progress, 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, 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 large models with the same performance as the backpropagation algorithm widely used in deep learning today. However, this challenge has been rapidly changing and a diverse ecosystem of training techniques provides clues for how PNNs may one day be used to create both more efficient and larger-scale realizations of present-scale AI models.
- 3 September 2025
- Published by Springer Nature in Nature
- Vol. 645 (8079) , 53-61
Top Cited PapersTraining of physical neural networks
Physical neural networks (PNNs) are a class of neural-like networks that make use of analogue physical systems to perform computations. Although at present confined to small-scale laboratory demonstrations, PNNs could one day transform how artificial intelligence (AI) calculations are performed. Could we train AI models many orders of magnitude larger than present ones? Could we perform model inference locally and privately on edge devices? Research over the past few years has shown that the answer to these questions is probably “yes, with enough research”. Because PNNs can make use of analogue physical computations more directly, flexibly and opportunistically than traditional computing hardware, they could change what is possible and practical for AI systems. To do this, however, will require notable progress, 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, 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 large models with the same performance as the backpropagation algorithm widely used in deep learning today. However, this challenge has been rapidly changing and a diverse ecosystem of training techniques provides clues for how PNNs may one day be used to create both more efficient and larger-scale realizations of present-scale AI models.
- Top Cited Papers
- 15 December 2023
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 382 (6676) , 1297-1303
Backpropagation-free training of deep physical neural networks
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 is unsuitable for physical implementation. In this work, we propose a simple deep neural network architecture augmented by a physical local learning (PhyLL) algorithm, which enables supervised and unsupervised training of deep physical neural networks without detailed knowledge of the nonlinear physical layer’s properties. We trained diverse wave-based physical neural networks in vowel and image classification experiments, showcasing the universality of our approach. Our method shows advantages over other hardware-aware training schemes by improving training speed, enhancing robustness, and reducing power consumption by eliminating the need for system modeling and thus decreasing digital computation.
- 15 December 2023
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 382 (6676) , 1297-1303
Top Cited PapersBackpropagation-free training of deep physical neural networks
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 is unsuitable for physical implementation. In this work, we propose a simple deep neural network architecture augmented by a physical local learning (PhyLL) algorithm, which enables supervised and unsupervised training of deep physical neural networks without detailed knowledge of the nonlinear physical layer’s properties. We trained diverse wave-based physical neural networks in vowel and image classification experiments, showcasing the universality of our approach. Our method shows advantages over other hardware-aware training schemes by improving training speed, enhancing robustness, and reducing power consumption by eliminating the need for system modeling and thus decreasing digital computation.
- 5 March 2021
- Published by SPIE-Intl Soc Optical Eng
- p. 13
Multimode fiber projector with neural networks
SPIE Digital Library Proceedings
- 5 March 2021
- Published by SPIE-Intl Soc Optical Eng
- p. 13
Multimode fiber projector with neural networks
SPIE Digital Library Proceedings
- 1 January 2021
- Published by Optica Publishing Group
- p. Th5B.1
Learning to See and Compute through Multimode Fibers
We propose a computational method for controlling the output of a multimode fiber using machine learning. Arbitrary images can be projected with amplitude-only calibration (no phase measurement) and fidelities on par with conventional full-measurement methods. We also show the reverse, meaning that multimode fibers can be used as a computational tool that harnesses spatiotemporal nonlinear effects to perform end to end learning tasks with unprecedented speed and low power consumption.
- 1 January 2021
- Published by Optica Publishing Group
- p. Th5B.1
Learning to See and Compute through Multimode Fibers
We propose a computational method for controlling the output of a multimode fiber using machine learning. Arbitrary images can be projected with amplitude-only calibration (no phase measurement) and fidelities on par with conventional full-measurement methods. We also show the reverse, meaning that multimode fibers can be used as a computational tool that harnesses spatiotemporal nonlinear effects to perform end to end learning tasks with unprecedented speed and low power consumption.
- 15 July 2020
- Published by Springer Nature in Nature Machine Intelligence
- Vol. 2 (7) , 403-410
Actor neural networks for the robust control of partially measured nonlinear systems showcased for image propagation through diffuse media
- 15 July 2020
- Published by Springer Nature in Nature Machine Intelligence
- Vol. 2 (7) , 403-410
Actor neural networks for the robust control of partially measured nonlinear systems showcased for image propagation through diffuse media
- 16 June 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Antennas and Propagation
- Vol. 69 (1) , 379-386
Wide-Band/Angle Blazed Dual-Mode Metallic Groove Gratings
A simple approach to acquire wide-band/angle blazing operation in one-dimensional metallic gratings is investigated by means of mode-matching and equivalent circuit analysis. The gratings are single-groove perfectly conducting gratings that support two propagating guided modes (opposed to typical single mode cases). It is shown under what conditions, one can achieve blazing over a wide range of frequencies and angles. Most importantly, it is identified what governs blazing, in particular the true matching condition that needs to be satisfied, and how the structures’ multiple resonances play role in achieving such matching and tailoring the bandwidth. Parameters of the proposed equivalent circuit model for the case of the transverse magnetic (TM) polarization are analytically derived. The accuracy of the models is verified through comparison against results of fullwave simulations. The procedure to achieve wide-band/angle blazing performance is delineated and the design parameters are explicitly given. It is shown that this structure can strongly transfer the power of the transverse magnetic (TM) polarized incident wave to the -1-st diffraction order with a fractional bandwidth of ≈ 50% at working frequency f0 = 10 GHz and for {10dB specular reflection loss, thus realizing a broadband structure. From the circuit model, an insightful discussion is presented at various mechanisms at play, when Bragg and off- Bragg blazing occurs and fully justifies the behavior of the device. The proposed technique opens up new vistas in a wide range of applications such as spectroscopy, Littrow cavities, beam splitters, refractive index biosensors, and frequency scanned antenna reflectors.
- 16 June 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Antennas and Propagation
- Vol. 69 (1) , 379-386
Wide-Band/Angle Blazed Dual-Mode Metallic Groove Gratings
A simple approach to acquire wide-band/angle blazing operation in one-dimensional metallic gratings is investigated by means of mode-matching and equivalent circuit analysis. The gratings are single-groove perfectly conducting gratings that support two propagating guided modes (opposed to typical single mode cases). It is shown under what conditions, one can achieve blazing over a wide range of frequencies and angles. Most importantly, it is identified what governs blazing, in particular the true matching condition that needs to be satisfied, and how the structures’ multiple resonances play role in achieving such matching and tailoring the bandwidth. Parameters of the proposed equivalent circuit model for the case of the transverse magnetic (TM) polarization are analytically derived. The accuracy of the models is verified through comparison against results of fullwave simulations. The procedure to achieve wide-band/angle blazing performance is delineated and the design parameters are explicitly given. It is shown that this structure can strongly transfer the power of the transverse magnetic (TM) polarized incident wave to the -1-st diffraction order with a fractional bandwidth of ≈ 50% at working frequency f0 = 10 GHz and for {10dB specular reflection loss, thus realizing a broadband structure. From the circuit model, an insightful discussion is presented at various mechanisms at play, when Bragg and off- Bragg blazing occurs and fully justifies the behavior of the device. The proposed technique opens up new vistas in a wide range of applications such as spectroscopy, Littrow cavities, beam splitters, refractive index biosensors, and frequency scanned antenna reflectors.
- Open Access
- 30 May 2020
- Published by MDPI AG in Applied Sciences
- Vol. 10 (11) , 3816
Deep Learning-Based Image Classification through a Multimode Fiber in the Presence of Wavelength Drift
Deep neural networks (DNNs) are employed to recover information after its propagation through a multimode fiber (MMF) in the presence of wavelength drift. The intensity distribution of the speckle patterns generated at the output of an MMF when an input wavefront propagates along its length is highly sensitive to wavelength changes. We use a tunable laser to implement a wavelength drift with a controlled bandwidth, aiming to estimate the DNN’s performance in different cases and identify the limitations. We find that when the DNNs are trained with a dataset which includes the noise induced by wavelength changes, successful classification of a speckle pattern can be performed even for a large wavelength bandwidth drift. A single training step is found to be sufficient for high classification accuracy, removing the need for time-consuming recalibration at each wavelength.
- 30 May 2020
- Published by MDPI AG in Applied Sciences
- Vol. 10 (11) , 3816
Open AccessDeep Learning-Based Image Classification through a Multimode Fiber in the Presence of Wavelength Drift
Deep neural networks (DNNs) are employed to recover information after its propagation through a multimode fiber (MMF) in the presence of wavelength drift. The intensity distribution of the speckle patterns generated at the output of an MMF when an input wavefront propagates along its length is highly sensitive to wavelength changes. We use a tunable laser to implement a wavelength drift with a controlled bandwidth, aiming to estimate the DNN’s performance in different cases and identify the limitations. We find that when the DNNs are trained with a dataset which includes the noise induced by wavelength changes, successful classification of a speckle pattern can be performed even for a large wavelength bandwidth drift. A single training step is found to be sufficient for high classification accuracy, removing the need for time-consuming recalibration at each wavelength.
- Open Access
Competing Neural Networks for Robust Control of Nonlinear Systems
The output of physical systems is often accessible by measurements such as the 3D position of a robotic arm actuated by many actuators or the speckle patterns formed by shining the spot of a laser pointer on a wall. The selection of the input of such a system (actuators and the shape of the laser spot respectively) to obtain a desired output is difficult because it is an ill-posed problem i.e. there are multiple inputs yielding the same output. In this paper, we propose an approach that provides a robust solution to this dilemma for any physical system. We show that it is possible to find the appropriate input of a system that results in a desired output, despite the input-output relation being nonlinear and\or with incomplete measurements of the systems variables. We showcase our approach using an extremely ill-posed problem in imaging. We demonstrate the projection of arbitrary shapes through a multimode fiber (MMF) when a sample of intensity-only measurements are taken at the output. We show image projection fidelity as high as ~90 %, which is on par with the gold standard methods which characterize the system fully by phase and amplitude measurements. The generality as well as simplicity of the proposed approach provides a new way of target-oriented control in real-world applications.
Open AccessCompeting Neural Networks for Robust Control of Nonlinear Systems
The output of physical systems is often accessible by measurements such as the 3D position of a robotic arm actuated by many actuators or the speckle patterns formed by shining the spot of a laser pointer on a wall. The selection of the input of such a system (actuators and the shape of the laser spot respectively) to obtain a desired output is difficult because it is an ill-posed problem i.e. there are multiple inputs yielding the same output. In this paper, we propose an approach that provides a robust solution to this dilemma for any physical system. We show that it is possible to find the appropriate input of a system that results in a desired output, despite the input-output relation being nonlinear and\or with incomplete measurements of the systems variables. We showcase our approach using an extremely ill-posed problem in imaging. We demonstrate the projection of arbitrary shapes through a multimode fiber (MMF) when a sample of intensity-only measurements are taken at the output. We show image projection fidelity as high as ~90 %, which is on par with the gold standard methods which characterize the system fully by phase and amplitude measurements. The generality as well as simplicity of the proposed approach provides a new way of target-oriented control in real-world applications.
- 1 January 2020
- Published by Optica Publishing Group
- p. HTu5B.3
Multimode fiber projection with machine learning
We propose an imaging method for controlling the output of scattering media such as multimode fibers using machine learning. Arbitrary images can be projected with amplitude-only calibration (no phase measurement) and fidelities on par with conventional full-measurement methods.
- 1 January 2020
- Published by Optica Publishing Group
- p. HTu5B.3
Multimode fiber projection with machine learning
We propose an imaging method for controlling the output of scattering media such as multimode fibers using machine learning. Arbitrary images can be projected with amplitude-only calibration (no phase measurement) and fidelities on par with conventional full-measurement methods.