110 results about "Artificial neuron" patented technology

The invention relates to the technical field of body-equipped intelligent robots, and discloses a multi-mode body-equipped intelligent robot control device, which comprises a brain-like decision model, a brain-like control strategy, a robot body and a sensing system which are connected in sequence, in the brain-like decision model part, an artificial neuron is adopted to construct a network to make a decision on behaviors of the robot, and an upper-layer control command is generated; the brain-like control strategy part is used for converting an upper-layer control command into a bottom-layer control command according to a specific application scene; the robot body executes a specified task and generates an action according to the received bottom layer control command; and the sensing system obtains the environment state change condition caused by the action of the robot body and updates the state information as the input of the brain-like decision model part at the next moment, so that a new round of control loop is started. According to the brain-like control strategy, a man-machine coordination motion control method is adopted for cooperative control, and a better interaction force control effect and higher position following precision are achieved for control over the robot body.

A device, system, and method is provided for storing a sparse neural network. A plurality of weights of the sparse neural network may be obtained. Each weight may represent a unique connection between a pair of a plurality of artificial neurons in different layers of a plurality of neuron layers. A minority of pairs of neurons in adjacent neuron layers are connected in the sparse neural network. Each of the plurality of weights of the sparse neural network may be stored with an association to a unique index. The unique index may uniquely identify a pair of artificial neurons that have a connection represented by the weight. Only non-zero weights may be stored that represent connections between pairs of neurons (and zero weights may not be stored that represent no connections between pairs of neurons).

The invention belongs to the related field of artificial nerve morphology, and discloses a compact artificial neuron based on TaOx multifunctional devices, the artificial neuron is composed of two TaOx multifunctional devices with the same structure, and the two TaOx multifunctional devices are connected in series back to back; wherein when voltage in a first direction is applied to the TaOx multifunctional device, the TaOx multifunctional device shows a charge trapping and releasing mechanism and is used for providing simulated volatile characteristics; and when the voltage in the second direction is applied, a conductive wire mechanism is shown and is used for providing digital volatile characteristics. The invention further discloses a structural design of the TaOx multifunctional device. According to the invention, the required neuron function can be realized by only relying on the back-to-back connection of two same TaOx multifunctional devices without additionally introducing capacitors, resistors and other devices, the circuit structure is compact and convenient to process and manufacture, and the problems of low integration level, high process difficulty and the like of the existing artificial neuron are effectively improved. Therefore, large-scale integration of the artificial neural network is facilitated.

The invention is particularly directed to a hardware system (1) designed for implementing an artificial neural network (ANN). The hardware system substantially includes a neural processing device (15) (e.g., involving a crossbar array structure), one or more lookup table circuits (17), and one or more processing units (18). The neural processing device is configured to implement M artificial neurons, where M > = 1. The lookup table circuit is configured to implement a lookup table (LUT). The system also includes M'processing units, where M > = M '> = 1. Each processing unit is connected by at least one neuron so as to be able to access a first value output by each connected neuron. In addition, each processing unit is connected to the LUT circuitry so as to efficiently access parameter values of the parameter set from the LUT. Finally, each processing unit is configured to output a second value corresponding to a value of a mathematical function with the first value as an argument. The mathematical function is additionally determined by a set of parameters whose parameter values are accessed in operation from the LUT by each processing unit. That is, the mathematical function is defined (and thus determined) by a set of parameters, the values of which can be efficiently retrieved from a hardware-implemented LUT. This results in a significant acceleration of the calculation of the function output beyond the acceleration that may have been achieved within the neural processing devices and processing units themselves. Thus, neuron outputs can be processed more efficiently before passing to the next neuron layer. The invention is also directed to a method of operating such a hardware system.

The invention is notably directed to a hardware system (1) designed to implement an artificial neural network (ANN). The hardware system basically includes a neural processing apparatus (15), e.g., involving as crossbar array structure, one or more lookup table circuits (17), and one or more processing units (18). The neural processing apparatus is configured to implement M artificial neurons, where M≥1. The lookup table circuits are configured to implement a lookup table (LUT). The system further includes M′ processing units, where M≥M′≥1. Each processing unit is connected by at least one neuron, in order to be able to access a first value outputted by each connected neuron. In addition, each processing unit is connected to a LUT circuit, in order to efficiently access parameter values of a set of parameters from the LUT. Finally, each processing unit is configured to output a second value, corresponding to a value of a mathematical function taking said first value as argument. The mathematical function is otherwise determined by the set of parameters, the parameter values of which are accessed by each processing unit from the LUT, in operation. I.e., the mathematical function is defined (and thus determined) by a set of parameters, the values of which are efficiently retrieved from the hardware-implemented LUT. This results in a substantial acceleration of the computations of the function outputs, beyond the acceleration that may already be achieved within the neural processing apparatus and the processing units themselves. As a result, the neuron outputs can be more efficiently processed, prior to being passed to a next neuron layer. The invention is further directed to a method of operating such a hardware system.

The invention provides a ferromagnetic artificial neuron film and application thereof. The ferromagnetic artificial neuron film comprises a transparent substrate and a ferromagnetic heterojunction constructed on the transparent substrate, the ferromagnetic heterojunction comprises a ferromagnetic alloy layer, a heavy metal layer and an anti-oxidation layer. Nonlinear spin oscillation of the ferromagnetic artificial neuron film is excited through femtosecond laser, terahertz radiation with intensity nonlinear response is directly generated, and the ferromagnetic artificial neuron film serves as an artificial neuron of a space optical neural network. The activation function response meeting the performance requirement of the artificial neuron can be effectively generated without introducing the regulation and control of an external magnetic field, and a better artificial neuron function facing the optical space neural network is realized.

Artificial neural networks (ANNs) are computing systems inspired by the human brain by learning to perform tasks by considering examples. These ANNs are typically created by connecting several layers of artificial neurons using connections, where each artificial neuron is connected to every other artificial neuron either directly or indirectly to create fully connected layers within the ANN. By substituting ternary matrices for one or more fully connected layers within the ANN, a complexity and resource usage of the ANN may be reduced, while improving the performance of the ANN.

Artificial neural networks (ANNs) are computing systems inspired by the human brain by learning to perform tasks by considering examples. These ANNs are typically created by connecting several layers of artificial neurons using connections, where each artificial neuron is connected to every other artificial neuron either directly or indirectly to create fully connected layers within the ANN. By substituting ternary matrices for one or more fully connected layers within the ANN, a complexity and resource usage of the ANN may be reduced, while improving the performance of the ANN.

The application provides a neuron-like transistor device, which comprises a substrate, a buried oxygen layer and a channel layer which are sequentially stacked on the substrate, a gate oxide layer and a gate electrode on the channel layer, and a source electrode and a drain electrode which are respectively located on both sides of the channel layer. The gate electrode is used for applying a front gate voltage as an input signal, the drain electrode is used for applying a constant input current, the voltage of the drain electrode is used as an output signal, and the input signal is changed in a step-by-step manner from small to large. The neuron nonlinear activation of a single device is realized, and it is possible to build a complete and programmable artificial neuron by using a single or few transistors. The neuron-like transistor device has a simple structure, and a manufacturing process which is suitable for a CMOS manufacturing process. The measurement method of the neuron-like transistor device is simple, and an electrode does not need to be additionally introduced, so that the simulation of the neuron characteristics can be realized without additional device design, the integration density of the transistors in a chip is improved, and the neuron-like transistor device is suitable for the integration of a large-scale circuit.

An artificial neuron includes a first capacitive node of application of a membrane potential of the neuron. A first transistor is configured to discharge the first capacitive node. A second capacitive node is driven according to the membrane potential and delivers a potential for controlling the first transistor. A second transistor is configured to discharge the second capacitive node. The second transistor is controlled according to a potential present at the second capacitive node.

The present invention discloses a high-frequency modulation oscillator device for composite memristor and liquid pool electrochemical reaction, which relates to multiple fields such as electronic component technology, neuromorphic devices, and electronic information processing. The high-frequency modulation oscillator device comprises a DC power supply, a first electrode, a redox liquid pool, a second electrode, and a memristor. The conductive state transition is achieved through a threshold switching effect. By virtue of the series connection of the redox liquid pool and the memristor working circuit, a solid-liquid interface working environment is constructed. Driven by DC voltages of different amplitudes, the device can achieve stable oscillation in a wide frequency range from a few tenths of a hertz to several kilohertz. The device is suitable for multiple application scenarios such as simulating artificial neuron behavior, generating broadband pulse trains, and processing digital signals.

A method for applying an error correction code (ECC) to a memory, performed by an ECC-application device in a system semiconductor based on an artificial neural network, comprises dividing memory cells into a plurality of groups of memory cell arrangements corresponding to the first to nth frames based on word lines, where n is a natural number greater than or equal to 2, performing an ECC write or read operation by applying the ECC to a memory cell arrangement of one group corresponding to each of the first to nth frames, and performing a general write or read operation without applying the ECC to the memory cell arrangement of the remaining groups except for the group corresponding to each of the first to nth frames.

The invention relates to a method for testing an artificial neural network (ANN), e.g., in the development of software applications for, in particular safety-relevant, functions, e.g., in vehicles, such as automated and / or autonomous vehicles. exhibiting: - Providing a, in particular trained, model (M) of the artificial neural network (ANN), - Defining dropout masks (Mi) to assess the uncertainty (U) of the artificial neural network (ANN) in a decision, - Determining at least one test result, in particular deterministic and preferably reproducible, e.g. in the form of at least one key performance indicator (KPI), of the artificial neural network (ANN).

According to one aspect, a computer-implemented method for compiling a first trained artificial neural network is proposed. This first trained artificial neural network comprises at least one succession (SUCC_2) of layers, including a depth-convolutional layer, followed by a saturation-rectified linear unit layer, and then a two-dimensional convolutional layer. The method comprises: - an equalization (13) between layers, - a replacement (15) of the saturation-rectified linear unit layer with a per-channel adaptive trimming layer to obtain an artificial neural network with a modified topology, - a tensor quantization (16) of the layers of the artificial neural network with a modified topology, - a compilation (17) of the quantized artificial neural network with a modified topology. Figure for the abstract: Figure 3

A neuromorphic computing device according to the present disclosure may have a plurality of artificial neurons connected to a synapse array, and each of the plurality of artificial neurons may include: a ferroelectric transistor having a gate connected to a first node, and connected between a power terminal and a second node configured to output an output spike, a first input transistor having a gate configured to receive a first input spike, and connected between a first input power terminal and the first node, a second input transistor having a gate configured to receive a second input spike, and connected between a second input power terminal and the first node, an adjustment transistor having a gate receiving an adjustment voltage and connected between the second node and ground terminal, and a reset transistor having a gate receiving a reset voltage, and connected between the first node and ground terminal.

The application belongs to the field of artificial neuron devices, in order to solve the problems of single function, complex manufacturing process and inability to adapt to mass production of the existing artificial neuron devices, the application provides a preparation method of a micro multifunctional artificial neuron device based on a two-dimensional WSe2 transistor, comprising the following steps: preparing a single-layer / few-layer selenium disulfide thin channel; evaporating Se / Au electrodes to manufacture a P-type WSe2 transistor; evaporating Bi / Au electrodes to manufacture an N-type WSe2 transistor and connecting by adopting a CMOS structure; the application also tests the electrical performance of the prepared device and the input and output curves of the activation function thereof, simulates various indexes and obtains the accuracy of the simulation results, and proves that the application can be applied to realize the activation function and the artificial synapse function.

A method and an apparatus for dual-neuron large-scale intelligent optical computing are disclosed. The method includes: constructing an optoelectronic computing unit, and obtaining an optical neuron by modeling the optoelectronic computing unit; abstracting the optical neuron as an artificial neuron based on a mathematical abstraction manner; constructing an optoelectronic neural network by connecting a plurality of artificial neurons, and performing a global optimization on network parameters of all the artificial neurons by applying end-to-end gradient descent on the optoelectronic neural network; optimizing physical parameters of the optical neuron based on the artificial neuron after the global optimization; and computing the optoelectronic neural network based on the optimized parameters of the optical neuron.

The invention discloses a dual-mode artificial neuron memristor and a manufacturing method thereof. The memristor device sequentially comprises a bottom electrode, a resistive layer and a top electrode from bottom to top, the bottom electrode is a P-type highly-doped pure silicon wafer, the resistive layer is an oxide thin film layer, and the top electrode is a thermally-evaporated silver electrode. The resistance change mechanism of the memristor is mainly the connection and disconnection of the silver conductive wire, and different switching behaviors can be displayed under different limiting currents, so that the memristor can switch a volatile mode and a non-volatile mode by adjusting the limiting current of soft breakdown, and if the limiting current Icc during soft breakdown is relatively large, the memristor can be switched to be in a non-volatile mode or a non-volatile mode by adjusting the limiting current Icc during soft breakdown. The device shows stable non-volatility; if the Icc is small, the silver conductive wire is fragile and easy to break, so that the device shows stable volatility. In addition, the memristor device can also simulate the integral ignition behavior of the IF model.

Synaptic reuse allows for a plurality of artificial neurons to be associated with corresponding pluralities of artificial synapses and variable gain amplifiers, to thereby use less space and fewer components to implement and less power to operate than neurons having dedicated paths for each input source. To reduce the likelihood of signal collision, and allow for the independent control and interpretation of input spikes, a router is configured to connect input sources to each of the plurality of artificial neurons in conjunction with a gain configuration controller that is configured to set a gain on each of the plurality of variable gain amplifiers based on a time division schema and an identity of an input source transmitting a spike during a given time.

The present invention relates to a method for generating training data for an artificial neural network for classifying anomalies in image data, in particular for classifying anomalies in image data from an environmental sensor of a motor vehicle.According to the invention, the method comprises the following steps: • Generating visual blocks on a substantially horizontal plane to replicate anomaly structures; capturing a labeling image using a visual sensor, which includes the replicated anomaly structures; creating labels for the replicated anomaly structures based on the captured labeling image; displaying environmental sensor data from a motor vehicle's environmental sensor using a display device; capturing a superimposed image using the visual sensor, which includes the replicated anomaly structures and the displayed environmental sensor data; creating a training dataset with labeled anomalies based on the superimposed image and the created labels for the replicated anomaly structures.Furthermore, the invention relates to a device configured to carry out the corresponding method, as well as a corresponding computer program.

The present application relates to the field of artificial neuromorphics, and provides an artificial neuron circuit based on a volatile threshold switching device. A membrane potential generation circuit is used for generating different membrane potentials under different input currents and different external reset voltages; a slow variable generation circuit is used for generating a slow variable; a membrane potential reset circuit comprises a volatile threshold switching device and an external reset voltage excitation source; the external reset voltage excitation source is used for providing different external reset voltages, so that the membrane potential and the slow variable are changed, and a spiking behavior of biological neurons is simulated; and the volatile threshold switching device is used for realizing a membrane potential reset function. Compared with traditional CMOS-circuit-based neurons, the neuron circuit provided by the present application requires a greatly reduced number of transistors and reduced hardware and power consumption overhead, has the characteristics of simple structure, high flexibility and rich functions, and is beneficial to large-scale integration in a hardware pulse neural network.

The ReLU artificial neurons play a key nonlinear activation function role in the deep artificial neural network. The ReLU artificial neuron is realized on a hardware circuit and is based on a special Pt / Ag / Al2O3 / HfO2 / Ag-NIs (5nm) / Pt threshold switching device embedded with a nano island, and the device has extremely low threshold voltage, low dispersity and high tolerance. The ReLU artificial neuron circuit is mainly composed of a threshold switching device and a parallel capacitor Cm, and a diode D1 and an adjusting resistor Rs are connected in series in front of the threshold switching device and the parallel capacitor Cm. An RC filtering unit composed of a capacitor Cf and a resistor Rf and a resistor Ro used for output detection are added to the output end of the circuit. Furthermore, a resistor Rd is added to be connected with the diode D1 in parallel, so that the improved Leaky ReLU artificial neuron circuit is formed. The artificial neuron provided by the invention has a great application prospect in constructing a brain-like artificial neural network.

The embodiment of the application discloses a kind of bridge artificial neuron unit composition's bionic hearing positioning system based on memristor circuit, it is related to neural morphic computing and artificial intelligence hardware field.In the present application: bridge artificial neuron unit includes the first memristor analog circuit connected with first capacitor and first load resistance, and the second memristor analog circuit connected with second capacitor and second load resistance, which are connected by first bridge capacitor, and coupled with external circuit by second bridge capacitor.The unit can simulate the pulse emission and refractory behavior of biological neuron.A plurality of the unit can be connected in series to build a one-way pulse propagation chain, forming a pulse propagation network.Based on the network, a coincidence detector can detect the small time difference of multiple input signals and be applied to a bionic hearing system to achieve sound source positioning.The present application has simple structure, high degree of modularity, and can realize high-reliability analog pulse propagation and high-precision coincidence detection.

The present invention provides a method for calculating decision variables. A dummy layer is added at an input layer of a trained neural network predictive model. The dummy layer includes a plurality of artificial neurons respectively connected to a corresponding input terminal of the input layer for the trained predictive model by a newly established link. The input value of each artificial neuron is set to 1, the bias value of the activation function is set to 0, and the output of the activation function is set to 1 when the input of the activation function is 1. The initial weight value of the newly established link is selected and set, and the weight values can be considered as decision variables, wherein the weight values can have ranges or other inter-conditional restrictions. The optimal solution is obtained using the optimizer built in a general machine-learning platform when the parameters of the trained predictive model are frozen, and only the weight values of the newly established links are adjusted. The training objective is set so that the output of the parameter predictive model matches the desired target result. At the end of the training, the weight values of the newly established link new are the feasible input decision variables. This invention allows the effective use of the general machine learning platforms and the built-in methods to find the optimal input parameters that achieve the expected results.

The present invention is synthetic neurons and networks with feedback quantum tunneling memristors designed to mimic biological neurons. The memristor hardware is formed from a 4.2 nm thick layer of atomic layer deposited ionic hafnium oxide and niobium metal and inserted in the positive and negative feedback of an analog spiking oscillator. When operated at room / warm temperatures, these memories have memristive properties and enable the artificial neuron circuits to produce adaptive spiking behavior. When neural networks are formed of the synthetic neurons, pronounced hybrid chaotic / non-chaotic modes with increased complexity are observed and itinerant behavior emerges. Cryogenic cooling of the memristors into the superconducting Josephson tunneling regime at 8.1 K reveals the influence of quantum control effects.

The application is a kind of bionic nerve circuit model based on artificial synapse device. The circuit model is composed of n multi-stage connections of synapse-LIF artificial neuron devices. The composition of the synapse-LIF artificial neuron device includes synapse-like device and cytoplasm-like device. The synapse-like device includes artificial synapse device, voltage dividing resistor and voltage control switch. The cytoplasm-like device includes one comparator. Through different combinations of the four devices, namely excitatory synapse-like device, inhibitory synapse-like device, excitatory cytoplasm-like device and inhibitory cytoplasm-like device, the devices of six circuits, including feedforward excitation circuit, feedforward inhibition circuit, feedback inhibition circuit, lateral inhibition circuit, disinhibition circuit and mutual inhibition circuit, are formed. The application realizes the functions of perception afferent, information integration and motor control in the motor nervous system through the circuit model group, which is helpful for the repair or reconstruction of damaged motor conduction circuit.

A system including a supply valve disposed between a chamber and a pressure source, a discharge valve disposed between the chamber and an external atmosphere, a first control unit, and a second control unit. The first control unit coupled with the supply valve by a first switch and with the discharge valve by a second switch. The first control unit outputting signals to the first and second switches to control the supply and discharge valves. The second control unit coupled with the discharge valve by a third switch and a fourth switch, the second control unit outputting signals to the third and fourth switches to control the supply and discharge valves. The first control unit may include a first microcontroller to control the signals of the first control unit using an artificial intelligence (AI) neural network having artificial neurons arranged in layers and connected with each other by connections.

The invention relates to a method for generating training data for learning an artificial neural network of a driver assistance system for a motor vehicle (1). The invention also relates to a computer system (100) implementing such a method. Figure for the abstract: 1

A hybrid neuromorphic computing device is provided, in which artificial neurons include light-emitting devices that provide weighted sums of inputs as light output. The output is detected by a photodetector and converted to an electrical output. Each neuron may receive output from one or more other neurons as initial input. Interconnects between neurons may be optical, electrical, or a combination thereof. The neurons also may provide imaging sensor and / or display capabilities.

A system and method for computing in memory with artificial neurons. According to an embodiment of the present disclosure, there is provided a system, including: a computer-readable memory; a neuron processing element communicatively connected to the computer-readable memory, the neuron processing element including: a plurality of configurable processing circuits each having a plurality of outputs and a plurality of inputs; and a network connecting one or more of the outputs of the configurable processing circuits to one or more of the inputs of the configurable processing circuits, each of the configurable processing circuits including: an artificial neuron having a plurality of inputs; and a register connected to the inputs of the artificial neuron.