392 results about "Neuron network" patented technology

Apparatus and methods for detecting salient features. In one implementation, an image processing apparatus utilizes latency coding and a spiking neuron network to encode image brightness into spike latency. The spike latency is compared to a saliency window in order to detect early responding neurons. Salient features of the image are associated with the early responding neurons. A dedicated inhibitory neuron receives salient feature indication and provides inhibitory signal to the remaining neurons within the network. The inhibition signal reduces probability of responses by the remaining neurons thereby facilitating salient feature detection within the image by the network. Salient feature detection can be used for example for image compression, background removal and content distribution.

Adaptive controller apparatus of a plant may be implemented. The controller may comprise an encoder block and a control block. The encoder may utilize basis function kernel expansion technique to encode an arbitrary combination of inputs into spike output. The controller may comprise spiking neuron network operable according to reinforcement learning process. The network may receive the encoder output via a plurality of plastic connections. The process may be configured to adaptively modify connection weights in order to maximize process performance, associated with a target outcome. The relevant features of the input may be identified and used for enabling the controlled plant to achieve the target outcome.

Efficient updates of connections in artificial neuron networks may be implemented. A framework may be used to describe the connections using a linear synaptic dynamic process, characterized by stable equilibrium. The state of neurons and synapses within the network may be updated, based on inputs and outputs to/from neurons. In some implementations, the updates may be implemented at regular time intervals. In one or more implementations, the updates may be implemented on-demand, based on the network activity (e.g., neuron output and/or input) so as to further reduce computational load associated with the synaptic updates. The connection updates may be decomposed into multiple event-dependent connection change components that may be used to describe connection plasticity change due to neuron input. Using event-dependent connection change components, connection updates may be executed on per neuron basis, as opposed to per-connection basis.

Apparatus and methods for conditional plasticity in a neural network. In one approach, conditional plasticity mechanism is configured to select alternate plasticity rules when performing connection updates. The selection mechanism is adapted based on a comparison of actual connection efficiency and target efficiency. For instance, when actual efficiency is below the target value, the STDP rule may be modulated to increase long term potentiation. Similarly, when actual efficiency is above the target value, the STDP rule may be modulated to increase long term connection depression. The conditional plasticity mechanism dynamically adjusts connection efficacy, and prevents uncontrolled increase of connection weights, thereby improving network operation when processing information of a varying nature.

Generalized state-dependent learning framework in artificial neuron networks may be implemented. A framework may be used to describe plasticity updates of neuron connections based on connection state term and neuron state term. The state connections within the network may be updated based on inputs and outputs to/from neurons. The input connections of a neuron may be updated using connection traces comprising a time-history of inputs provided via the connections. Weights of the connections may be updated and connection state may be time varying. The updated weights may be determined using a rate of change of the trace and a term comprising a product of a per-neuron contribution and a per-connection contribution configured to account for the state time-dependency. Using event-dependent connection change components, connection updates may be executed on per neuron basis, as opposed to per-connection basis.

Apparatus and methods for modulated plasticity in a spiking neuron network. A plasticity mechanism may be configured for example based on a similarity measure between post-synaptic activities of two or more neurons that may be receiving the same feed-forward input. The similarity measure may comprise a dynamically determined cross-correlogram between the output spike trains of two neurons. An a priori configured similarity measure may be used during network operation in order to update efficacy of inhibitory connections between neighboring neurons. Correlated output activity may cause one neuron to inhibit output generation by another neuron thereby hindering responses by multiple neurons to the same input stimuli. The inhibition may be based on an increased efficacy of inhibitory lateral connection. The inhibition may comprise modulation of the pre synaptic portion the plasticity rule based on efficacies of feed-forward connection and inhibitory connections and a statistical parameter associated with the post-synaptic rule.

State-dependent supervised learning framework in artificial neuron networks may be implemented. A framework may be used to describe plasticity updates of neuron connections based on a connection state term and a neuron state term. Connection states may be updated based on inputs and outputs to and/or from neurons. The input connections of a neuron may be updated using input traces comprising a time-history of inputs provided via the connection. Weight of the connection may be updated and connection state may be time varying. The updated weights may be determined using a rate of change of the input trace and a term comprising a product of a per-neuron contribution and a per-connection contribution configured to account for the state time-dependency. Using event-dependent connection change components, connection updates may be executed on a per neuron basis, as opposed to a per-connection basis.

The invention discloses a memristor-based neuron circuit, comprising a synaptic circuit, a neuron activation function circuit and a synaptic weight control circuit. In the synaptic circuit, a memristor, under the control of four MOS tubes, changes the memristor value to simulate the change of the synaptic weight in the neuron network. The designed neuron synaptic circuit is capable of being directly connected with a digital logic electrical level so as to achieve convenient and real-time adjustment to the synaptic weight and through the use of the feature that the output voltage of an operational amplifier is restricted by the power supply voltage, the neuron circuit activation function can be realized as a saturated linear function. The neuron synaptic weight change circuit can utilize the existing CMOS micro-controller and at the same time, the micro-controller can be loaded by the neuron network algorithm to change the synaptic weight to realize corresponding functions. According to the invention, a plurality of neuron circuits could be connected into a large-scale neuron network for complicated functions such as mode identification, signal processing, associated memory and non-linear mapping, etc.

Apparatus and methods for learning in response to temporally-proximate features. In one implementation, an image processing apparatus utilizes bi-modal spike timing dependent plasticity in a spiking neuron network. Based on a response by the neuron to a frame of input, the bi-modal plasticity mechanism is used to depress synaptic connections delivering the present input frame and to potentiate synaptic connections delivering previous and/or subsequent frames of input. The depression of near-contemporaneous input prevents the creation of a positive feedback loop and provides a mechanism for network response normalization.

Adaptive controller apparatus of a plant may be implemented. The controller may comprise an encoder block and a control block. The encoder may utilize basis function kernel expansion technique to encode an arbitrary combination of inputs into spike output. The controller may comprise spiking neuron network operable according to reinforcement learning process. The network may receive the encoder output via a plurality of plastic connections. The process may be configured to adaptively modify connection weights in order to maximize process performance, associated with a target outcome. The relevant features of the input may be identified and used for enabling the controlled plant to achieve the target outcome. The stochasticity of the learning process may be modulated. Stochasticity may be increased during initial stage of learning in order to encourage exploration. During subsequent controller operation, stochasticity may be reduced to reduce energy use by the controller.

Apparatus and methods for an extensible robotic device with artificial intelligence and receptive to training controls. In one implementation, a modular robotic system that allows a user to fully select the architecture and capability set of their robotic device is disclosed. The user may add/remove modules as their respective functions are required/obviated. In addition, the artificial intelligence is based on a neuronal network (e.g., spiking neural network), and a behavioral control structure that allows a user to train a robotic device in manner conceptually similar to the mode in which one goes about training a domesticated animal such as a dog or cat (e.g., a positive/negative feedback training paradigm) is used. The trainable behavior control structure is based on the artificial neural network, which simulates the neural/synaptic activity of the brain of a living organism.

The invention discloses a system and method for predicting user behavior in wireless Internet. The system comprises a user behavior data acquisition module which is positioned on a client and is used for acquiring user behavior data within user running time and transmitting the user behavior data to a server, and a user behavior analyzing and predicting module which is positioned on a server side and is used for establishing a user behavior model and analyzing and predicting user behavior according to user behavior data acquired by the user behavior data acquisition module positioned on the client. The method comprises the following steps of: A, establishing a user behavior model; B, acquiring user behavior data within user running time; and C, and analyzing and predicting user behavior according to the acquired user behavior data. According to the technical scheme of invention, the integrity and effectiveness of data can be ensured; and behavior prediction is partitioned into group long-term behavior prediction and individual short-term behavior prediction, so that the correlation between user property and user behavior is fully mined. Comprehensive behavior prediction is performed by using a neural network, so that a prediction result is perfected and made accurate continuously.

The invention discloses a four-rotor aircraft control method based on PID neural network (PIDNN) control. First, dynamic equations of a four-rotor aircraft are established based on the Newton-Euler equations, and then, a nested controller is proposed. An inner loop decentralized controller is designed based on a PIDNN method to realize attitude control, and the outer loop adopts a classical PID control method. Next, online learning of a PIDNN controller is realized by an error back propagation method, the connection initial weight of PIDNN is determined based on the principle of PID control, and an appropriate learning step is selected according to the discrete Lyapunov theory to guarantee the convergence of the controller. Thus, the control problem of a four-rotor aircraft caused by high degree of nonlinearity, strong coupling and under-actuating capability, disturbance from the external environment in the process of flight and other factors can be effectively solved.

Described is a system for representing, storing, and reconstructing an input signal. The system constructs an index of unique polychronous groups (PCGs) from a spiking neuron network. Thereafter, a basis set of spike codes is generated from the unique PCGs. An input signal can then be received, with the input signal being spike encoded using the basis set of spike codes from the unique PCGs. The input signal can then be reconstructed by looking up in a reconstruction table, for each unique PCG in the basis set in temporal order according to firing times, anchor neurons. Using a neuron assignment table, an output location can be looked up for each anchor neuron to place a value based on the firing times of each unique PCG. Finally, the output locations of the anchor neurons can be compiled to reconstruct the input signal.

Apparatus and methods for contrast enhancement and feature identification. In one implementation, an image processing apparatus utilizes latency coding and a spiking neuron network to encode image brightness into spike latency. The spike latency is compared to a saliency window in order to detect early responding neurons. Salient features of the image are associated with the early responding neurons. An inhibitory neuron receives salient feature indication and provides inhibitory signal to the other neurons within an area of influence of the inhibitory neuron. The inhibition signal reduces probability of responses by the other neurons to stimulus that is proximate to the feature thereby increasing contrast within the encoded data. The contrast enhancement may facilitate feature identification within the image. Feature detection may be used for example for image compression, background removal and content distribution.

The invention provides an Izhikevich neural network synchronous discharging simulation platform based on an FPGA. The simulation platform comprises an FPGA neural network computing processor and an upper computer which are connected with each other. The FPGA neural network computing processor comprises an FPGA chip, an off-chip memorizer array and an Ethernet communication module, wherein the FPGA chip receives an upper computer control signal output by the off-chip memorizer array, and receives a presynaptic membrane potential signal output by the off-chip memorizer array. The upper computer is in communication with the FPGA chip and the off-chip memorizer array through a VB programming realization man-machine operating interface and the Ethernet communication module, and a neural network model is established on the FPGA chip through Verilog HDL language programming. The Izhikevich neural network synchronous discharging simulation platform has the advantages that the hardware modeling of the phenotype and physiological type neural network model is achieved through an animal-free experiment serving as a biological neural network on the basis of an FPGA neural network experiment platform conducting computation at a high speed, and the consistency with true biological nerve cells on the time scale can be achieved.

The invention relates to a monocular infrared image depth estimation method based on an optimized BP (Back Propagation) neural network model. The method comprises the following steps of: acquiring a monocular infrared image and a depth map to which the monocular infrared image corresponds; setting at least three feature regions with different scales for pixel points in the monocular infrared image; calculating feature vectors of the feature regions to which the pixel points in the monocular infrared image correspond; screening all the feature vectors by successively using stepwise linear regression and independent component analysis methods to obtain feature vectors conforming to depth information of the infrared image; constructing a depth training sample set by using the obtained feature vectors and the depth map to which the infrared image corresponds, and performing nonlinear fitting on the feature vectors in the set and depth values of the depth map by using a BP neuron network, and optimizing the BP neuron network through a genetic algorithm, and then constructing a depth model; and analyzing the monocular infrared image through the depth model to obtain a depth estimated value. By using the monocular infrared image depth estimation method based on the optimized BP neural network model, the depth information of the infrared image can be relatively accurately estimated.