Scalable Optical Learning Operator

Abstract

Today's heavy machine learning tasks are fueled by large datasets. Computing is performed with power hungry processors whose performance is ultimately limited by the data transfer to and from memory. Optics is one of the powerful means of communicating and processing information and there is intense current interest in optical information processing for realizing high-speed computations. Here we present and experimentally demonstrate an optical computing framework based on spatiotemporal effects in multimode fibers for a range of learning tasks from classifying COVID-19 X-ray lung images and speech recognition to predicting age from face images. The presented framework overcomes the energy scaling problem of existing systems without compromising speed. We leveraged simultaneous, linear, and nonlinear interaction of spatial modes as a computation engine. We numerically and experimentally showed the ability of the method to execute several different tasks with accuracy comparable to a digital implementation.

Main text: 18 pages, 7 figures. Supplementary material: 13 pages, 11 figures, 2 tables

References in corpus (13)

• Parallel convolution processing using an integrated photonic tensor core
• Spatiotemporal mode-locking in multimode fiber lasers
• Spatial beam self-cleaning in multimode fiber
• Observation of geometric parametric instability induced by the periodic spatial self-imaging of multimode waves
• Adaptive pumping for spectral control of random lasers
• Large-Scale Optical Reservoir Computing for Spatiotemporal Chaotic Systems Prediction
• Theory of neuromorphic computing by waves: machine learning by rogue waves, dispersive shocks, and solitons
• Single mode output by controlling the spatiotemporal nonlinearities in mode-locked femtosecond multimode fiber lasers
• Random Projections through multiple optical scattering: Approximating kernels at the speed of light
• Wavefront shaping in multimode fibers by transmission matrix engineering
• Classification of time-domain waveforms using a speckle-based optical reservoir computer
• Observation of Spatio-temporal Instability of Femtosecond Pulses in Normal Dispersion Multimode Graded-Index Fiber
• Gain As Configurable Disorder-Adaptive Pumping for Control of Multimode Fiber Amplifiers and Lasers

Cited by in corpus (25)

• The physics of optical computing
• Controlling light propagation in multimode fibers for imaging, spectroscopy and beyond
• Polarization Multiplexed Diffractive Computing: All-Optical Implementation of a Group of Linear Transformations Through a Polarization-Encoded Diffractive Network
• Nonlinear optical encoding enabled by recurrent linear scattering
• Massively Parallel Universal Linear Transformations using a Wavelength-Multiplexed Diffractive Optical Network
• Cascadable all-optical NAND gates using diffractive networks
• A photonic complex perceptron for ultrafast data processing
• Wave-based extreme deep learning based on non-linear time-Floquet entanglement
• Programmable wave-based analog computing machine: a metastructure that designs metastructures
• Broadband nonlinear modulation of incoherent light using a transparent optoelectronic neuron array
• Referenceless characterisation of complex media using physics-informed neural networks
• Multiplexed all-optical permutation operations using a reconfigurable diffractive optical network
• Predicting nonlinear optical scattering with physics-driven neural networks
• An optoacoustic field-programmable perceptron for recurrent neural networks
• Spectral Shaping in a Multimode Fiber by All-fiber Modulation
• Nonlinear Inference Capacity of Fiber-Optical Extreme Learning Machines
• Deep learning-assisted imaging through stationary scattering media
• Coupling to multi-mode waveguides with space-time shaped free-space pulses
• High-performance real-world optical computing trained by in situ gradient-based model-free optimization
• Image classification using collective optical modes of an array of nanolasers
• Optical neuromorphic computing via temporal up-sampling and trainable encoding on a telecom device platform
• High-speed coherent photonic random-access memory in long-lasting sound waves
• Arbitrary control over multimode wave propagation for machine learning
• Limits of nonlinear and dispersive fiber propagation for an optical fiber-based extreme learning machine
• Integrated Random Projection and Dimensionality Reduction by Propagating Light in Photonic Lattices