Abstract: Disclosed is a neural network-based inference method and apparatus. The neural network-based inference method includes compressing a matrix comprising processing elements corresponding to an operation of a neural network, balancing workloads related to the operation by reordering the compressed matrix based on the workloads, and performing inference based on the reordered matrix.

Abstract: A processor-implemented neural network method includes: determining a reference sample among sequential input samples to be processed by a neural network, the neural network comprising an input layer, one or more hidden layers, and an output layer; performing an inference process of obtaining an output activation of the output layer based on operations in the hidden layers corresponding to the reference sample input to the input layer; determining layer contraction parameters for determining an affine transformation relationship between the input layer and the output layer, for approximation of the inference process; and performing inference on one or more other sequential input samples among the sequential input samples using affine transformation based on the layer contraction parameters determined with respect to the reference sample.

Abstract: A processor-implemented neural network method includes: determining a reference sample among sequential input samples to be processed by a neural network, the neural network comprising an input layer, one or more hidden layers, and an output layer; performing an inference process of obtaining an output activation of the output layer based on operations in the hidden layers corresponding to the reference sample input to the input layer; determining layer contraction parameters for determining an affine transformation relationship between the input layer and the output layer, for approximation of the inference process; and performing inference on one or more other sequential input samples among the sequential input samples using affine transformation based on the layer contraction parameters determined with respect to the reference sample.

Abstract: A neurocomputational electrochemical sensing device (1) is proposed for predicting properties of a substance. The device (1) comprises: a plurality of electrochemical sensors constituting a sensor array (3), the sensors being sensitive to sensed attributes advantageous to predict a set of properties of interest of the substance, each sensor being configured to output a sensor output signal indicative of a sensor response of the respective sensor to measurable changes in the sensed attributes of the substance; a readout circuit (5) for biasing the sensors and for conditioning the sensor output signals into readout circuit output signals to facilitate further processing of the sensor responses; and an artificial neural network processor (7) for processing the readout circuit output signals, the processor (7) comprising neurons interconnected by synapses, the processor (7) being configured to output a set of processor output signals whose signal values are indicative of the properties to predict.

Abstract: A sensor transformation attention network (STAN) model including sensors configured to collect input signals, attention modules configured to calculate attention scores of feature vectors corresponding to the input signals, a merge module configured to calculate attention values of the attention scores, and generate a merged transformation vector based on the attention values and the feature vectors, and a task-specific module configured to classify the merged transformation vector is provided.

Abstract: A method for operating an artificial neuron and an apparatus for performing the method are provided. The artificial neuron may calculate a change amount of an activation based on an input signal received via an input synapse, determine whether an event occurs in response to the calculated change amount of the activation, and transmit, to an output synapse, an output signal that corresponds to the event in response to an occurrence of the event.

Abstract: A method for operating an artificial neuron and an apparatus for performing the method are provided. The artificial neuron may calculate a change amount of an activation based on an input signal received via an input synapse, determine whether an event occurs in response to the calculated change amount of the activation, and transmit, to an output synapse, an output signal that corresponds to the event in response to an occurrence of the event.

Abstract: A processor-implemented neural network method includes acquiring connection weight of an analog neural network (ANN) node of a pre-trained ANN; and determining, a firing rate of a spiking neural network (SNN) node of an SNN, corresponding to the ANN node, based on an activation of the ANN node which is determined based on the connection weight. and the firing rate is also determined based on information indicating a timing at which the SNN node initially fires.

Abstract: A modular artificial neural sensing system includes a hierarchical network of neural sensing units including a neuromimetic sensor array of artificial sensory synapses and sensory neurons for receiving physicochemical sensed signals and for outputting sensor output signals. An artificial neural network processor is adapted for processing the sensor output signals and includes processor neurons interconnected by processor synapses forming first connections and second connections. The processor outputs processor output signals. A first sensor interface feeds processed or unprocessed sensed signals into the processor. A second sensor interface receives output predicted signals from other neural sensing units and feeds processed or unprocessed output predicted signals into the processor. A signal decoder decodes the processor output signals and outputs decoder output signals.

Abstract: A computer-implemented method is used to adapt a first artificial neural network for data classification tasks. The first artificial neural network is characterized by a first number of first weight parameters, and includes a set of first network layers. The method includes freezing at least some of the first weight parameters of the first neural network to obtain frozen first weight parameters and duplicating the frozen first weight parameters to obtain duplicated first weight parameters. A second artificial neural network is applied to the duplicated first weight parameters to obtain modulated first weight parameters. The second artificial neural network is characterized by a second number of second weight parameters, the second number being smaller than the first number. The frozen first weight parameters are replaced in the first neural network with the modulated first weight parameters to obtain a modulated first artificial neural network adapted for a data classification task.

Abstract: A sensor transformation attention network (STAN) model including sensors, attention modules, a merge module and a task-specific module is provided. The attention modules calculate attention scores of feature vectors corresponding to the input signals collected by the sensors. The merge module calculates attention values of the attention scores, and generates a merged transformation vector based on the attention values and the feature vectors. The task-specific module classifies the merged transformation vector.

Abstract: Disclosed is a neural network-based inference method and apparatus. The neural network-based inference method includes compressing a matrix comprising processing elements corresponding to an operation of a neural network, balancing workloads related to the operation by reordering the compressed matrix based on the workloads, and performing inference based on the reordered matrix.

Abstract: A method for operating an artificial neuron and an apparatus for performing the method are provided. The artificial neuron may calculate a change amount of an activation based on an input signal received via an input synapse, determine whether an event occurs in response to the calculated change amount of the activation, and transmit, to an output synapse, an output signal that corresponds to the event in response to an occurrence of the event.

Abstract: A method for operating an artificial neuron and an apparatus for performing the method are provided. The artificial neuron may calculate a change amount of an activation based on an input signal received via an input synapse, determine whether an event occurs in response to the calculated change amount of the activation, and transmit, to an output synapse, an output signal that corresponds to the event in response to an occurrence of the event.

Abstract: A method for operating an artificial neuron and an apparatus for performing the method are provided. The artificial neuron may calculate a change amount of an activation based on an input signal received via an input synapse, determine whether an event occurs in response to the calculated change amount of the activation, and transmit, to an output synapse, an output signal that corresponds to the event in response to an occurrence of the event.

Abstract: A processor-implemented neural network method includes: determining a reference sample among sequential input samples to be processed by a neural network, the neural network comprising an input layer, one or more hidden layers, and an output layer; performing an inference process of obtaining an output activation of the output layer based on operations in the hidden layers corresponding to the reference sample input to the input layer; determining layer contraction parameters for determining an affine transformation relationship between the input layer and the output layer, for approximation of the inference process; and performing inference on one or more other sequential input samples among the sequential input samples using affine transformation based on the layer contraction parameters determined with respect to the reference sample.

Abstract: A band-pass filter is described comprising a first first-order filter stage comprising a first resistor characterized by a first impedance and connected to a first node, referred to as a filter input node, and, through a second node to a first reactive component connected to a third node, the first impedance being such that a first current therethrough is dependent on the difference between the voltages at the first and second nodes; and a second first-order filter stage comprising a second resistor characterized by a second impedance and connected to the second node, and, through a fourth node, to a second reactive component connected to a fifth node. The second impedance is such that a second current therethrough is dependent on the negative of the sum of the voltages at the second and fourth nodes. The band-pass filter further comprises summing means for summing the voltages at the second and fourth nodes to output a voltage at a sixth node.

Abstract: A recurrent neural network including an input layer, a hidden layer, and an output layer, wherein the hidden layer includes hybrid memory cell units, each of the hybrid memory cell units including a first memory cells of a first type, the first memory cells being configured to remember a first cell state value fed back to each of gates to determine a degree to which each of the gates is open or closed, and configured to continue to update the first cell state value, and a second memory cells of a second type, each second memory cell of the second memory cells including a first time gate configured to control a second cell state value of the second memory cell based on phase signals of an oscillatory frequency, and a second time gate configured to control an output value of the second memory cell based on the phase signals, and each second memory cell of the second memory cells being configured to remember the second cell state value.

Abstract: A processor-implemented neural network method includes acquiring connection weight of an analog neural network (ANN) node of a pre-trained ANN; and determining, a firing rate of a spiking neural network (SNN) node of an SNN, corresponding to the ANN node, based on an activation of the ANN node which is determined based on the connection weight. and the firing rate is also determined based on information indicating a timing at which the SNN node initially fires.

Abstract: A band-pass filter is described comprising a first first-order filter stage comprising a first resistor characterised by a first impedance and connected to a first node, referred to as a filter input node, and, through a second node to a first reactive component connected to a third node, the first impedance being such that a first current therethrough is dependent on the difference between the voltages at the first and second nodes; and a second first-order filter stage comprising a second resistor characterised by a second impedance and connected to the second node, and, through a fourth node, to a second reactive component connected to a fifth node. The second impedance is such that a second current therethrough is dependent on the negative of the sum of the voltages at the second and fourth nodes. The band-pass filter further comprises summing means for summing the voltages at the second and fourth nodes to output a voltage at a sixth node.