Deep learning performance is limited by computing power, which is in turn limited by energy consumption. Optics can make neural networks faster and more efficient, but current schemes suffer from limited connectivity and the large footprint of low-loss nanophotonic devices. Our optical neural network architecture addresses these problems using homodyne detection and optical data fan-out. It is scalable to large networks without sacrificing speed or consuming too much energy. It can perform inference and training and work with both fully connected and convolutional neural-network layers. In our architecture, each neural network layer operates on inputs and weights encoded onto optical pulse amplitudes. A homodyne detector computes the vector product of the inputs and weights. The nonlinear activation function is performed electronically on the output of this linear homodyne detection step. Optical modulators combine the outputs from the nonlinear activation function and encode them onto optical pulses input into the next layer.

Deep neural networks (DNNs) have become very popular in many areas, especially classification and prediction. However, as the number of neurons in the DNN increases to solve more complex problems, the DNN becomes limited by the latency and power consumption of existing hardware. A scalable, ultra-low latency photonic tensor processor can compute DNN layer outputs in a single shot. The processor includes free-space optics that perform passive optical copying and distribution of an input vector and integrated optoelectronics that implement passive weighting and the nonlinearity. An example of this processor classified the MNIST handwritten digit dataset (with an accuracy of 94%, which is close to the 96% ground truth accuracy). The processor can be scaled to perform near-exascale computing before hitting its fundamental throughput limit, which is set by the maximum optical bandwidth before significant loss of classification accuracy (determined experimentally).

An object classification system for classifying objects is described. The system comprises an imaging region adapted for irradiating an object of interest, an arrayed detector, and a mixing unit configured for mixing the irradiation stemming from the object of interest by reflecting or scattering on average at least three times the irradiation after its interaction with the object of interest and prior to said detection.

Methods and apparatus for training a matrix-based differentiable program using a photonics-based processor. The matrix-based differentiable program includes at least one matrix-valued variable associated with a matrix of values in a Euclidean vector space. The method comprises configuring components of the photonics-based processor to represent the matrix of values as an angular representation, processing, using the components of the photonics-based processor, training data to compute an error vector, determining in parallel, at least some gradients of parameters of the angular representation, wherein the determining is based on the error vector and a current input training vector, and updating the matrix of values by updating the angular representation based on the determined gradients.

The ability to rapidly identify symmetry and anti-symmetry is an essential attribute of intelligence. Symmetry perception is a central process in human vision and may be key to human 3D visualization. While previous work in understanding neuron symmetry perception has concentrated on the neuron as an integrator, the invention provides the coincidence detecting property of the spiking neuron can be used to reveal symmetry density in spatial data. A synchronized symmetry-identifying spiking artificial neural network enables layering and feedback in the network. The network of the invention can identify symmetry density between sets of data and present a digital logic implementation demonstrating an 8×8 leaky-integrate-and-fire symmetry detector in a field-programmable gate array. The efficiency of spiking neural networks can be harnessed to rapidly identify symmetry in spatial data with applications in image processing, 3D computer vision, and robotics.

The present disclosure discloses a nonlinear all-optical deep-learning system and method with multistage space-frequency domain modulation. The system includes an optical input module, configured to convert input information to optical information, a multistage space-frequency domain modulation module, configured to perform multistage space-frequency domain modulation on the optical information generated by the optical input module so as to generate modulated optical information, and an information acquisition module, configured to transform the modulated optical information onto a Fourier plane or an image plane, and to acquire the transformed optical information so as to generate processed optical information.

Entity models are used to evaluate potential risk of entities, either individually or in groups, in order to evaluate suspiciousness within an enterprise network. These individual or aggregated risk assessments can be used to adjust the security policy for compute instances within the enterprise network. A security policy may specify security settings such as network speed, filtering levels, network isolation, levels of privilege, and the like.

Embodiments of the present disclosure are directed toward techniques and configurations for optical couplers comprising a first optical waveguide and a second optical waveguide coupled to form a 2×2 optical unitary matrix to receive a respective first input optical signal and a second input optical signal. In embodiments the first optical waveguide and second optical waveguide form arms that converge alongside each other to direct the first input optical signal and the second input optical signal along a path that integrates a plurality of tunable phase shifters to transform the first input optical signal or the second input optical signal into a first output optical signal and second output optical signal to be output from the 2×2 optical unitary matrix. Additional embodiments may be described and claimed.

A diffractive optical neural network device includes a plurality of diffractive substrate layers arranged in an optical path. The substrate layers are formed with physical features across surfaces thereof that collectively define a trained mapping function between an optical input and an optical output. A plurality of groups of optical sensors are configured to sense and detect the optical output, wherein each group of optical sensors has at least one optical sensor configured to capture a positive signal from the optical output and at least one optical sensor configured to capture a negative signal from the optical output. Circuitry and/or computer software receives signals or data from the optical sensors and identifies a group of optical sensors in which a normalized differential signal calculated from the positive and negative optical sensors within each group is the largest or the smallest of among all the groups.

A photonic computing system is presented. The system comprises an arrangement of multiple photonic processing units having input and output ports, each of the photonic processing units comprising an array of photonic guiding units configured to define propagation conditions for multiple light fields associated with one or more optical processing tasks. The system also comprises a plurality of optical connectors, each of the optical connectors performing light field to light field coupling between the input and output ports of the photonic processing units, thereby providing a network of communicating processing units. The photonic computing system can be configured as a module enabling its housing in a network rack.