All Optical Neural Network with Nonlinear Activation Functions

Abstract

Artificial neural networks (ANNs) have now been widely used for industry applications and also played more important roles in fundamental researches. Although most ANN hardware systems are electronically based, optical implementation is particularly attractive because of its intrinsic parallelism and low energy consumption. Here, we propose and demonstrate fully-functioned all optical neural networks (AONNs), in which linear operations are programmed by spatial light modulators and Fourier lenses, and optical nonlinear activation functions are realized with electromagnetically induced transparency in laser-cooled atoms. Moreover, all the errors from different optical neurons here are independent, thus the AONN could scale up to a larger system size with final error still maintaining in a similar level of a single neuron. We confirm its capability and feasibility in machine learning by successfully classifying the order and disorder phases of a typical statistic Ising model. The demonstrated AONN scheme can be used to construct various ANNs of different architectures with the intrinsic parallel computation at the speed of light.

arXiv comment6 pages, 4 figures

References in corpus (18)

• Deep Learning with Coherent Nanophotonic Circuits
• All-Optical Machine Learning Using Diffractive Deep Neural Networks
• Solving the Quantum Many-Body Problem with Artificial Neural Networks
• Machine learning phases of matter
• Optoelectronic Reservoir Computing
• All-optical Reservoir Computing
• Training of photonic neural networks through in situ backpropagation
• Reinforcement Learning in a large scale photonic Recurrent Neural Network
• Machine Learning Topological States
• Efficient quantum memory for single photon polarization qubits
• Quantum Loop Topography for Machine Learning
• Self-Learning Monte Carlo Method
• Accelerate Monte Carlo Simulations with Restricted Boltzmann Machines
• Machine learning meets quantum physics
• Unsupervised Learning of Frustrated Classical Spin Models I: Principle Component Analysis
• Self-Learning Monte Carlo Method in Fermion Systems
• A Compact, Transportable, Microchip-Based System for High Repetition Rate Production of Bose-Einstein Condensates
• A Compact Microchip-Based Atomic Clock Based on Ultracold Trapped Rb Atoms

Cited by in corpus (63)

• Photonics for artificial intelligence and neuromorphic computing
• Single-chip photonic deep neural network for instantaneous image classification
• Large-scale neuromorphic optoelectronic computing with a reconfigurable diffractive processing unit
• Intelligent Computing: The Latest Advances, Challenges and Future
• Image sensing with multilayer, nonlinear optical neural networks
• Machine-Learning-Assisted Metasurface Design for High-Efficiency Thermal Emitter Optimization
• Spectrally-Encoded Single-Pixel Machine Vision Using Diffractive Networks
• Neuromorphic computing using wavelength-division multiplexing
• Deep Learning with Coherent VCSEL Neural Networks
• Large-Scale Optical Reservoir Computing for Spatiotemporal Chaotic Systems Prediction
• Polarization Multiplexed Diffractive Computing: All-Optical Implementation of a Group of Linear Transformations Through a Polarization-Encoded Diffractive Network
• Photonics for Neuromorphic Computing: Fundamentals, Devices, and Opportunities
• Computational inverse design for cascaded systems of metasurface optics
• Photonic extreme learning machine by free-space optical propagation
• Free-space optical neural network based on thermal atomic nonlinearity
• Silicon Photonic Architecture for Training Deep Neural Networks with Direct Feedback Alignment
• Massively Parallel Universal Linear Transformations using a Wavelength-Multiplexed Diffractive Optical Network
• Noise-enhanced spatial-photonic Ising machine
• Universal Linear Intensity Transformations Using Spatially-Incoherent Diffractive Processors
• All-Optical Spiking Neuron Based On Passive Micro-Resonator
• An Electro-Photonic System for Accelerating Deep Neural Networks
• Large-scale Ising Emulation with Four-Body Interaction and All-to-All Connection
• Polarized deep diffractive neural network for classification, generation, multiplexing and de-multiplexing of orbital angular momentum modes
• Universal metasurfaces for complete linear control of coherent light transmission
• Scalability of all-optical neural networks based on spatial light modulators
• Ultrahigh-fidelity spatial mode quantum gates in high-dimensional space by diffractive deep neural networks
• Photonic architecture for reinforcement learning
• Energy-efficient neural network inference with microcavity exciton-polaritons
• Wave-based extreme deep learning based on non-linear time-Floquet entanglement
• Broadband nonlinear modulation of incoherent light using a transparent optoelectronic neuron array
• Robust and Efficient Single-Pixel Image Classificationwith Nonlinear Optics
• Silicon photonic-electronic neural network for fibre nonlinearity compensation
• Single-Pixel Pattern Recognition with Coherent Nonlinear Optics
• An optoacoustic field-programmable perceptron for recurrent neural networks
• All-Optical Image Identification with Programmable Matrix Transformation
• Realistic quantum photonic neural networks
• Optical convolutional neural network with atomic nonlinearity
• Decision-making and control with diffractive optical networks
• Neural networks based on ultrafast time-delayed effects in exciton-polaritons
• Express Wavenet -- a low parameter optical neural network with random shift wavelet pattern
• Single-Shot Matrix-Matrix Multiplication Optical Tensor Processor for Deep Learning
• Backpropagation through nonlinear units for all-optical training of neural networks
• High-speed coherent photonic random-access memory in long-lasting sound waves
• Optical neural network architecture for deep learning with the temporal synthetic dimension
• An optical diffractive deep neural network with multiple frequency-channels
• Dark-state and loss-induced phenomena in the quantum-optical regime of $\Lambda$-type three-level systems
• Limits of nonlinear and dispersive fiber propagation for an optical fiber-based extreme learning machine
• Engineering nonlinear activation functions for all-optical neural networks via quantum interference
• Image amplification in a self-imaging degenerate optical cavity
• Quantum optical classifier with superexponential speedup
• Adiabatic evolution on a spatial-photonic Ising machine
• Nonlinear unitary circuits for photonic neural networks
• An All-optical Proteretic Switch Using Semiconductor Ring Lasers
• Performance evaluation of an integrated photonic convolutional neural network based on delay buffering and wavelength division multiplexing
• Unitary Learning for Deep Diffractive Neural Network
• Sub-quadratic scalable approximate linear converter using multi-plane light conversion with low-entropy mode mixers
• All-optical neural network quantum state tomography
• A single metagrating metastructure for wave-based parallel analog computing
• A Diffractive Neural Network with Weight-Noise-Injection Training
• Forming complex neurons by four-wave mixing in a Bose-Einstein condensate
• Modeling and benchmarking quantum optical neurons for efficient neural computation
• Steady states of $\Lambda$-type three-level systems excited by quantum light in lossy cavities
• All-Optical Synthesis of an Arbitrary Linear Transformation Using Diffractive Surfaces