50 results about "Synaptic junction" patented technology

Apparatus and methods for feedback in a spiking neural network. In one approach, spiking neurons receive sensory stimulus and context signal that correspond to the same context. When the stimulus provides sufficient excitation, neurons generate response. Context connections are adjusted according to inverse spike-timing dependent plasticity. When the context signal precedes the post synaptic spike, context synaptic connections are depressed. Conversely, whenever the context signal follows the post synaptic spike, the connections are potentiated. The inverse STDP connection adjustment ensures precise control of feedback-induced firing, eliminates runaway positive feedback loops, enables self-stabilizing network operation. In another aspect of the invention, the connection adjustment methodology facilitates robust context switching when processing visual information. When a context (such an object) becomes intermittently absent, prior context connection potentiation enables firing for a period of time. If the object remains absent, the connection becomes depressed thereby preventing further firing.

A circuit element of a multi-dimensional dynamic adaptive neural network array (DANNA) may comprise a neuron / synapse select input functional to select the circuit element to function as one of a neuron and a synapse. In one embodiment of a DANNA array of such circuit elements, (wherein a circuit element may be digital), a destination neuron may be connected to a first neuron by a first synapse in one dimension a second destination neuron may be connected to the first neuron by a second synapse in a second dimension to form linked columns and rows of neuron / synapse circuit elements. In one embodiment, the rows and columns of circuit elements have read registers that are linked together by signal lines and clocked and controlled so as to output columnar data to an output register when a neuron / synapse data value is stored in the read register.

Apparatus and methods for feedback in a spiking neural network. In one approach, spiking neurons receive sensory stimulus and context signal that correspond to the same context. When the stimulus provides sufficient excitation, neurons generate response. Context connections are adjusted according to inverse spike-timing dependent plasticity. When the context signal precedes the post synaptic spike, context synaptic connections are depressed. Conversely, whenever the context signal follows the post synaptic spike, the connections are potentiated. The inverse STDP connection adjustment ensures precise control of feedback-induced firing, eliminates runaway positive feedback loops, enables self-stabilizing network operation. In another aspect of the invention, the connection adjustment methodology facilitates robust context switching when processing visual information. When a context (such an object) becomes intermittently absent, prior context connection potentiation enables firing for a period of time. If the object remains absent, the connection becomes depressed thereby preventing further firing.

A method and apparatus for constructing a neuroscience-inspired dynamic architecture (NIDA) for an artificial neural network is disclosed. The method comprises constructing, in one embodiment, an artificial neural network embodiment in a multi-dimensional space in memory such that a neuron is connected by a synapse to another neuron. The neuron and the synapse each have parameters and have features of long-term potentiation and long-term depression. Furthermore, crossover and mutation are employed to select children of parents. Through learning, an initial network may evolve into a different network when NIDA is applied to solve different problems of control, anomaly detection and classification over selected time units. The apparatus comprises in one embodiment a computational neuroscience-inspired artificial neural network with at least one affective network coupled to receive input data from an environment and to output data to the environment.

Apparatus and methods for feedback in a spiking neural network. In one approach, spiking neurons receive sensory stimulus and context signal that correspond to the same context. When the stimulus provides sufficient excitation, neurons generate response. Context connections are adjusted according to inverse spike-timing dependent plasticity. When the context signal precedes the post synaptic spike, context synaptic connections are depressed. Conversely, whenever the context signal follows the post synaptic spike, the connections are potentiated. The inverse STDP connection adjustment ensures precise control of feedback-induced firing, eliminates runaway positive feedback loops, enables self-stabilizing network operation. In another aspect of the invention, the connection adjustment methodology facilitates robust context switching when processing visual information. When a context (such an object) becomes intermittently absent, prior context connection potentiation enables firing for a period of time. If the object remains absent, the connection becomes depressed thereby preventing further firing.

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.

A circuit element of a multi-dimensional dynamic adaptive neural network array (DANNA) may comprise a neuron / synapse select input functional to select the circuit element to function as one of a neuron and a synapse. In one embodiment of a DANNA array of such circuit elements, (wherein a circuit element or component thereof may be analog or digital), a destination neuron may be connected to a first neuron by a first synapse in one dimension, a second destination neuron may be connected to the first neuron by a second synapse in a second dimension and, optionally, a third destination neuron may be connected to the first neuron by a third synapse. The DANNA may thus form multiple levels of neuron and synapse circuit elements. In one embodiment, multiples of eight inputs may be selectively received by the circuit element selectively functioning as one of a neuron and a synapse. The dynamic adaptive neural network array (DANNA) may comprise a special purpose processor for performing one of a control, anomaly detection and classification application and may comprise a first structure connected to a neuroscience-inspired dynamic artificial neural network (NIDA), comprise substructures thereof or be combined with other neural networks.

The invention provides a method and device for inducing a conditioned neural response in a subject. The method comprises detecting spike activity in a first neural site in the subject; and delivering a stimulus pulse to a second neural site in the subject. The stimulus pulse is delivered within the time window for synaptic strengthening following the detecting of a spike. These two steps, detection and stimulation, are repeated continuously, typically for a day or two. The conditioned neural response is induced when a pattern of neural activity evoked by stimulation at the first neural site emulates a pattern of neural activity evoked by stimulation at the second neural site. The conditioned neural response persists for an extended period of time.

A digital circuit element of a two dimensional dynamic adaptive neural network array (DANNA) may comprise a neuron / synapse select input functional to select the digital circuit element to function as one of a neuron and a synapse. In one embodiment of a DANNA array of such digital circuit elements, a destination neuron may be connected to a first neuron by a first synapse in one dimension, a second destination neuron may be connected to the first neuron by a second synapse in a second dimension and, optionally, a third destination neuron may be connected to the first neuron by a third synapse thus forming multiple levels of neuron and synapse digital circuit elements. In one embodiment, multiples of eight inputs may be selectively received by the digital circuit element selectively functioning as one of a neuron and a synapse. The dynamic adaptive neural network array (DANNA) may implement long-term potentiation or depression to facilitate learning through the use of an affective system and random selection of input events.

In one approach, spiking neurons receive sensory stimulus and context signal that correspond to the same context. When the stimulus provides sufficient excitation, neurons generate response. Context connections are adjusted according to inverse spike-timing dependent plasticity. When the context signal precedes the post synaptic spike, context synaptic connections are depressed. Conversely, whenever the context signal follows the post synaptic spike, the connections are potentiated. The inverse STDP connection adjustment ensures precise control of feedback-induced firing, eliminates runaway positive feedback loops, and enables self-stabilizing network operation. In another aspect, the connection adjustment methodology facilitates robust context switching when processing visual information. When a context (such an object) becomes intermittently absent, prior context connection potentiation enables firing for a period of time. If the object remains absent, the connection becomes depressed thereby preventing further firing.

The invention discloses a neuromorphic visual target classification method based on an improved spiking neural network. The method comprises the following steps: S1, acquiring a neuromorphic visual target classification data set; s2, pulse event stream serialization aggregation: aggregating the space-time pulse event stream data in the data set into new event frame sequence data according to a set time resolution dt; s3, constructing an improved spiking neural network model: improving a synaptic connection mode of a leakage-accumulation-emission (LIF) spiking neuron in a time dimension, and constructing an improved spiking neural network based on an improved LIF neuron layer; s4, for a data set after pulse event flow serialization aggregation, randomly extracting samples from a sequence as input, and training and testing the constructed improved pulse neural network; and S5, storing the trained improved spiking neural network structure and network parameters. According to the invention, the network classification accuracy in target identification and classification of neuromorphic vision can be effectively improved.

The invention aims at the special requirement of the memristor as the electronic component of the processor, presents a neuron and neuron circuit based on the memristor of brain-like device, which canrealize the signal storage and processing similar to human neuron cell, and its single neuron cell can be extensively connected with hundreds of memristors, which provides a very feasible circuit design way for the large-scale use of memristors. In combination with other types of electronic devices such as CMOS tubes, a selector, a nanowire, and pulse design, the invention creatively solves the polysynaptic connection in biology of neuron cell designed by memristor as the core device of processor, forward stimulation, backward stimulation, nuclear storage, synaptic front-end, protruding back-end and other biological design issues, to achieve (forward and backward) signal processing within neurons and transmission between neurons, and build the corresponding neuronal cells and neural network circuits.

A neural network comprises a plurality of artificial neurons and a plurality of artificial synapses each input neuron being connected to each output neuron by way of an artificial synapse, the network being characterized in that each synapse consists of a first memristive device connected to a first input of an output neuron and of a second memristive device, mounted in opposition to said first device and connected to a second, complemented, input of said output neuron so that said output neuron integrates the difference between the currents originating from the first and second devices.

The invention relates to an embedded pile tip foundation detecting system, which comprises an embedded pile, a first sealing rod, a second sealing rod, a detecting rod and a processor. The embedded pile is of a tapered tubular structure; reinforced synaptic junctions are arranged on the outer walls of the embedded pile; a small port end of the embedded pile is formed by an arc-shaped plate; the first sealing rod is in threaded connection with the embedded pile, and is provided with a disassembly head part and an expansion head; a stirring plate is arranged on the expansion head; the second sealing rod comprises an outer rod and a piston rod; the outer rod is provided with a discharge hole and a loading chamber; the piston rod is inserted into the loading chamber from the other end of the outer rod; an ultrasonic probe is arranged on one end of the detecting rod, and a wireless signal transmitting and receiving device is arranged on the other end of the detecting rod; and the processor is in signal connection with the wireless signal transmitting and receiving device. According to the embedded pile tip foundation detecting system provided by the invention, after a cast-in-place pile is casted, the first sealing rod is pulled out, so that a cut-through channel is formed. The embedded pile tip foundation detecting system can be used for detecting without the need for carrying out core drilling operation, and is very convenient to detect and simple to operate. After the detecting is completed, the second sealing rod is inserted into the cast-in-place pile to seal the cut-through channel, so that the structural strength of the whole cast-in-place pile can be improved.

The present invention discloses a method to recover and restore dendritic and synaptic neuron connections that have been degraded or destroyed by neurodegenerative diseases. In the present invention tryptamines are used to induce neuro plasticity and restore both dendritic density and synaptic connections of neurons in the brain. In the preferred embodiment LSD given in micro doses can induce dendritic and synaptic genesis in neuronal networks and improve the quality of life of people with neurodegenerative diseases such as Alzheimer's, Huntington's, Multiple Sclerosis, Parkinson's and Frontotemporal dementia.

Neuromorphic processing apparatus is provided. The present invention may include a spiking neural network comprising a set of input spiking neurons each connected to each of a set of output spiking neurons via a respective synapse for storing a synaptic weight which is adjusted for that synapse in dependence on network operation in a learning mode of the apparatus, and each synapse is operable toprovide a post-synaptic signal, dependent on its synaptic weight, to its respective output neuron. The present invention may further include a pre-processor unit adapted to process input data, defining a pattern of data points, to produce a first set of input spike signals which encode values representing respective data points, and a second set of input spike signals which encode values complementary to respective said values representing data points, and to supply the input spike signals to respective predetermined input neurons of the network.

The invention discloses a path planning method and device based on an impulsive neural network. The method comprises steps of (1) dividing a map into a target region, a normal passing region and an obstacle region, modeling the map into the impulsive neural network, enabling each region to correspond to a neuron of the impulsive neural network, and enabling each neuron to be connected through twounidirectional synapses; (2) initializing neuron parameters and synaptic parameters corresponding to the target area; (3) at each moment, updating the membrane potential of each neuron according to the pulse condition of the incoming synapse and the self-charging current, sending a pulse signal when the membrane potential is greater than a potential threshold, transmitting the pulse signal to thenext neuron through the synapse, and returning the membrane potential to zero after the neuron transmits the pulse signal; (4) at each moment, updating a trace value of each synapse according to whether the pulse signal is transmitted or not, and updating the weight of the synapse according to the trace value; and (5) after the transmission of the pulse signal is finished, planning a path according to the weight value of the synapse to obtain a path result.

The invention discloses a simulation and training method of a pulse neural network with pulse moment offset, which comprises the following steps of: discretizing to update a neuron state quantity step by step, estimating whether pulse distribution is generated in a time step range and the moment offset thereof, calculating pulse input at the next moment according to synaptic connection and repeating the process to complete network simulation; and according to simulation result delay neuron state quantity, pulse emission quantity and time offset back propagation gradient error, updating network parameters. Errors in time step-by-step simulation calculation of the pulse neural network are reduced, and the simulation efficiency is improved under the same precision condition. And meanwhile, the pulse error is transmitted through two dimensions of the pulse emission amount and the time offset in the error back propagation algorithm, so that the training efficiency is improved.

The invention discloses a synaptic junction type handle structure comprising a handle, a supporting block, a base and a synaptic junction; the supporting block is fixed on the outer surface of the base; the handle is hinged to the supporting block; the front end of the handle is inclined towards the inner side, and the top end of the handle is hinged with the synaptic junction through a hinge shaft; the synaptic junction passes through the base, and the length of the synaptic junction is larger than the thickness of the base; the base is provided with a sliding chute matched with the synaptic junction; the top end of the handle is provided with an inclined surface; the distance from the hinge shaft to the supporting block is smaller than 1 / 3 of the distance from the tail of the handle to the supporting block; and the inclined surface on the top end of the handle is wider than the sliding chute. By using the synaptic junction type handle structure with the design, the handle automatically pushes the synaptic junction to stretch when a door body is opened, the door body is opened by utilizing point contact under the action of a lever, the top end of the handle props against the surface of the base when the handle is twisted to the maximum limiting position, and thus, the door body can be continued to be opened, and the time and the labor are saved.

The invention relates to a preparation method of an external wall dustproof coating, and belongs to the technical field of external wall coatings. The preparation method disclosed by the invention comprises the steps as follows: firstly, using malvaceae pollen with a micron-sized particle size as a raw material, and performing a reaction on the malvaceae pollen and 3,4-dihydroxyphenylalanine underthe action of catechol oxidase to obtain reaction filter residues; then performing a reaction on the reaction filter residues and a mixed saline solution of iron salt and sodium bicarbonate to obtaina self-made dustproof filler; then using organic silicon modified acrylic resin as a paint base stock; and finally, mixing the self-made dustproof filler with the organic silicon modified acrylic resin as well as other auxiliary materials to prepare paint, then brushing the paint on a wall surface, applying magnetic field processing to the brushed paint, and finally, naturally curing to form theexternal wall dustproof coating. The dustproof coating prepared with the preparation method disclosed by the invention is very excellent in dustproof effect; and as time goes by, a double-micro-synaptic-junction structure is generated on the surface of the coating, and thus, a contact angle is further increased, the dustproof effect is very excellent, and the coating has a wide application prospect.

The invention discloses an improved hippocampus-forehead cortex network space cognition method, which comprises the following steps: firstly, a robot explores an environment to collect information, and hippocampus position cells receive the environment information and form a class topology cognition map; cortical column neurons in the forehead cortex receive and integrate hippocampus information projection and environmental award stimulation signals to evaluate the class topology cognitive map, in the evaluation process, if the robot detects that the environment is dynamically changed, the emergency change module is excited, and if multiple targets exist in the environment, the award re-estimation module is excited; finally, stimulating signals are diffused in the model in the form of neural activity waves, after evaluation is completed, reward stimulating signals are diffused in the forehead cortex network, synaptic connection between cortical column neurons in the forehead cortex is changed according to an STDP rule, and a global vector field is formed; and finally, target-oriented space navigation of the robot is realized.

A spiking neural network includes a plurality of neurons implemented in respective circuits. Each neuron produces a continuous-valued membrane potential according to a Growth Transform bounded by an extrinsic energy constraint. The continuous-valued membrane potential is defined as a function of spiking current received from another neuron in the plurality of neurons, and a received electrical current stimulus. The spiking neural network includes a network energy function representing network energy consumed by the plurality of neurons and a neuromorphic framework. The neuromorphic framework minimizes network energy consumed by the plurality of neurons to determine the extrinsic energy constraint, models synaptic connections among the plurality of neurons as respective transconductances that regulate magnitude of spiking currents received from each of the plurality of neurons by each other of the plurality of neurons, and encodes the received electrical current stimulus in corresponding continuous-valued membrane potentials of the plurality of neurons.

The invention provides an in-vitro artificial reflex arc structure and a construction method and application thereof, and belongs to the technical field of tissue engineering. The in-vitro artificialreflex arc structure which is clear in neuron network structure and high in signal propagation reliability is constructed in vitro by combining the advantages of high structural precision of a micro-protein pattern method based on micro-contact printing and stable formation of cultured synaptic connection; neuron axon extension is restrained and guided by utilizing a one-way threshold unit micro-protein mask pattern, so that the threshold value and direction of signal propagation in the in-vitro artificial reflex arc structure are controllable; micro-contact printing is carried out by adoptinga micro-mechanical arm, so that the requirement on the operation method is reduced, and the protein pattern is stable in formation; and the characteristic that the one-way threshold unit micro-protein mask pattern allows growth of dense neuron cell populations is utilized, so that an additional neuroglial cell layer does not need to be added to serve as a nutrient supporting layer of neurons during culture, the culture requirement of the in-vitro artificial reflex arc structure is greatly reduced, and the culture process is simplified.

The invention discloses an automatic identification method and device for a whole brain range single neuron axis sudden synaptic junction. The method comprises the steps of carrying out the splittingof a neuron skeleton tree, obtaining a plurality of sub-segment fibers, and extracting the neighborhood image signals of the sub-segment fibers; carrying out the adaptive foreground segmentation on the neighborhood image signals of the sub-segment fibers, positioning the axonal expansion structures of the sub-segment fibers, and fusing the axonal expansion structures on all the sub-segment fibers;and classifying the axonal expansion structures by using a deep convolutional network, and removing the axonal expansion structures of the non-axonal sudden contact junctions in the axonal expansionstructures to obtain the final axonal sudden contact junctions. The device comprises an acquisition module and a classification module. According to the invention, the identification difficulties caused by the diverse shapes of the axonal synaptic junctions, adhesion of the axonal synaptic junctions, interference of false axonal synaptic junctions and the like can be overcome, all axonal synapticjunctions of a single nerve are quickly identified from a whole brain neuron morphological data set, and the accuracy of an identification result is kept while a whole brain oversized data set is quickly processed.

The invention discloses an image processing system architecture. The image processing system architecture comprises a pixel array, a synaptic array and a neuron circuit array. The synaptic array comprises multiple stages of synaptic arrays. The neuron circuit array comprises multiple stages of neuron circuit arrays, and any two neurons in each two stages of neuron circuit arrays are respectively connected by one synaptic in one synaptic array. According to the architecture, image processing does not need extra conversion between analog and digital signals, real-time parallel processing can berealized, and the processing speed is improved.

A convolutional neural network is provided comprising artificial neurons arranged in layers, each comprising output matrices. An output matrix comprises output neurons and is connected to an input matrix, comprising input neurons, by synapses associated with a convolution matrix comprising weight coefficients associated with the output neurons of an output matrix. Each synapse consists of a set of memristive devices storing a weight coefficient of the convolution matrix. In response to a change of the output value of an input neuron, the neural network dynamically associates each set of memristive devices with an output neuron connected to the input neuron. The neural network comprises accumulator(s) for each output neuron; to accumulate the values of the weight coefficients stored in the sets of memristive devices dynamically associated with the output neuron, the output value of the output neuron being determined from the value accumulated in the accumulator(s).

This invention provides nARIA-based assays for defining whether a subject has or is predisposed to developing a neoplastic disease. This invention further provides a nARIA-based method for maintaining synaptic connections between a neuron and a target cell.

The invention provides a robot behavior decision-making method and device based on a memory sequence playback mechanism, and the method comprises the steps: obtaining a target task and current environment information, wherein the current environment information comprises all obstacle information; constructing a neural network model, setting a dynamic adaptive competition mechanism in a development network, adding an advanced feature presentation layer to carry out online autonomous learning, and continuously updating synaptic connection weights of action neurons in an output layer based on a memory sequence playback mechanism until the network is stable; on the basis of the current environment information, obtaining a global optimal behavior decision through calculation of the neural network model, that is, the method can autonomously learn or identify information of irregular obstacles in a complex environment in online and offline states, so that the robot can adapt to a complex and changeable environment, the global optimal decision of the robot is realized, and the convergence speed of the neural network is improved.