2165 results about "Artificial neuronal network" patented technology

Methods and apparatuses for balancing computing workload via migrating computing tasks are disclosed. An artificial neural network (ANN) is trained based on the workload distribution over time for a host. The ANN predicts the workload for the host, and an indication may be sent to migrate at least one computing task away from the host. The indication is sent when the method is operating in a proactive mode and when the predicted workload is outside of a desired operating range. Some embodiments monitor the workload; and automatically switch the method to the proactive mode, when a difference between the monitored workload and the predicted workload is small. Other embodiments monitor the workload; and automatically switch the method to a reactive mode, when the monitored workload is outside of a failsafe operating range for the particular host.

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.

An enterprise system and method for maintaining and transitioning humans to a human-like self-reliant entity is presented. Said system including at least one a biological, biomechatronic, and mechatronic entity with a biological or artificial neural network to at least one transform or maintain. Embodiments are provided to assist in the transition of human between a biological state to a bio-mechatronic and mechatronic entity. Said entity's biological, biomechatronic, and mechatronic subsystems are configured to communicate and interact with one another in order for said enterprise system to manage, configure, maintain, and sustain said entity throughout the entity's life-cycle. Subsystem embodiments and components supported by the enterprise system are presented.

Apparatus and methods for universal node design implementing a universal learning rule in a mixed signal spiking neural network. In one implementation, at one instance, the node apparatus, operable according to the parameterized universal learning model, receives a mixture of analog and spiking inputs, and generates a spiking output based on the model parameter for that node that is selected by the parameterized model for that specific mix of inputs. At another instance, the same node receives a different mix of inputs, that also may comprise only analog or only spiking inputs and generates an analog output based on a different value of the node parameter that is selected by the model for the second mix of inputs. In another implementation, the node apparatus may change its output from analog to spiking responsive to a training input for the same inputs.

A computational method for identifying adhesin and adhesin-like proteins, said method comprising steps of computing the sequence-based attributes of a neural network software wherein the attributes are (i) amino acid frequencies, (ii) multiplet frequency, (iii) dipeptide frequencies, (iv),charge composition, and (v) hydrophobic composition, training the artificial neural Network (ANN) for each of the computed five attributes, and identifying the adhesin and adhesin-like proteins having probability of being an adhesin (P_ad) as ≧0.51; a computer system for performing the method; and genes and proteins encoding adhesin and adhesin-like proteins.

The invention relates to a method for automatic recognition and stage compression of medical image regions of interest based on an artificial neural network. Medical image files in a digital diagnostic system are generally larger, and due to the limitation by factors of bandwidth and the like, the transmission speed is low, the effect is not good, and the diagnosis quality is influenced. By the method, medical digital images are subjected to noise elimination, the tissue outline of a human body is recognized, tissue images are subjected to multiple times of overlay operation, image features of the regions of interest are strengthened, feature values are extracted, classification is performed by using an artificial neural network method, the regions of interest and corresponding levels are determined, and tagged image file format (TIFF) images are generated in different compression modes according to different levels of the regions of interest and non-regions of interest. By the method, the medical image files are greatly lessened, the transmission speed is increased, and effective necessary information used for diagnosis and treatment in the images is kept, so the method facilitates the reading of doctors, and can be applied to the digital diagnostic system and a remote medical system, and improve the diagnosis and treatment efficiency and effect.

The invention discloses a deeper layer network constructing method used for gender identification or age estimation based on human face. The method includes a step (101) dividing all training pictures into a plurality of groups; (102) extracting high layer features of a group of pictures based on a convolution neural network and thereby obtaining a first matrix composed of the high layer feature vectors, and extracting low layer and global features of the same group of the training images based on an artificial neural network and thereby obtaining a second matrix composed of the low layer feature vectors, obtaining a group of gender identification or age estimation results based on the extract first matrix, the second matrix and the defined judgment formula, wherein the values of a first weight matrix W1, a second weight matrix w2, an offset matrix b and an adjusting weight beta in the defined judgment formula are updated by utilizing an error back propagation algorithm and the final values of the parameters are obtained and the network construction is completed. Judgment of age and gender of a human face is performed based on the judgment formula determined according to the values of the parameters when the network construction is completed.

A method and apparatus for constructing one of a neuroscience-inspired artificial neural network and a neural network array comprises one of a neuroscience-inspired dynamic architecture, a dynamic artificial neural network array and a neural network array of electrodes associated with neural tissue such as a brain, the method and apparatus having a special purpose display processor. The special purpose display processor outputs a display over a period of selected reference time units to demonstrate a neural pathway from, for example, one or a plurality of input neurons through intermediate destination neurons to an output neuron in three dimensional space. The displayed neural network may comprise neurons and synapses in different colors and may be utilized, for example, to show the behavior of a neural network for classifying hand-written digits between values of 0 and 9 or recognizing vertical/horizontal lines in a grid image of lines.

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.

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 contamination level estimation method for high voltage insulators collects samples of naturally contaminated insulators and builds an image data set for the collected insulators. Flashover voltages of several insulators samples are measured. ESDD levels of the collected insulators are estimated. Images are input to image processing algorithms to extract representative features. The images are segmented. Transforming the image from RGB color space into grayscale model excludes the background from the image. Subsequently, the segmented images are transferred back to RGB color space model using matrix manipulation. Since contaminants on the insulator surface affect the color of the insulator, the segmented image is transformed from RGB to HSV color space which is used for extracting statistical and linear algebraic features from the hue image. A trained artificial neural network correlates the extracted features to the contamination levels enabling testing of other contaminated insulators.

A method includes determining object class probabilities of pixels in a first input image by examining the first input image in a forward propagation direction through layers of artificial neurons of an artificial neural network. The object class probabilities indicate likelihoods that the pixels represent different types of objects in the first input image. The method also includes selecting, for each of two or more of the pixels, an object class represented by the pixel by comparing the object class probabilities of the pixels with each other, determining an error associated with the object class that is selected for each pixel of the two or more pixels, determining one or more image perturbations by back-propagating the errors associated with the object classes selected for the pixels of the first input image through the layers of the neural network without modifying the neural network, and modifying a second input image by applying the one or more image perturbations to one or more of the first input image or the second input image prior to providing the second input image to the neural network for examination by the neurons in the neural network for automated object recognition in the second input image.

Systems and methods of using machine-learning in a cognitive radio to avoid a jammer are described. Smoothed power spectral density is used to detect activity in a sub-band and basic characteristics of different signals therein extracted. If unable to classify the signals as either a valid signal or a jammer using the basic characteristics, ANN-based classification with cumulants features of the signals is used. Multiple periods are used to train sensing and communications (S/C) polices to track and avoid a jammer using RL (e.g. Q learning). The ANN has input neurons of higher order cumulants of a sensing channel and a single output neuron. The S/C polices are coupled during training and communication using negative or decreasing rewards based on the time the sensing policy takes to determine jammer presence and that the cognitive radio is jammed. A feedback channel provides a new communications channel to a radio transmitting to the cognitive radio.

Methods and software utilizing artificial neural networks (ANNs) to estimate density and/or flow (speed) of objects in one or more scenes each captured in one or more images. In some embodiments, the ANNs and their training configured to provide reliable estimates despite one or more challenges that include but are not limited to, low-resolution images, low framerate image acquisition, high rates of object occlusions, large camera perspective, widely varying lighting conditions, and widely varying weather conditions. In some embodiments, fully convolutional networks (FCNs) are used in the ANNs. In some embodiments, a long short-term memory network (LSTM) is used with an FCN. In such embodiments, the LSTM can be connected to the FCN in a residual learning manner or in a direct connected manner. Also disclosed are methods of generating training images for training an ANN-based estimating algorithm that make training of the estimating algorithm less costly.

Provided in the invention is a hybrid computing system of an artificial neural network and an impulsive neural network. The hybrid computing system comprises a plurality of neural network computing units; one part of the neural network computing units is an artificial neural network computing unit being responsible for artificial neural network computing; and the other parts of the neural network computing units are impulsive neural network computing units responsible for impulsive neural network computing. The multiple basic computing units are mutually connected according a certain topological structure, thereby realizing a neural network computing function jointly. According to the hybrid computing system, the computing modes of the two kinds of neural networks containing the artificial neural network and the impulsive neural network are combined to realize real-time multi-mode or complex space-time signal calculation rapidly and guarantee the computing precision.

The present invention provides a method (30) for training an artificial neural network (NN). The method (30) includes the steps of: initialising the NN by selecting an output of the NN to be trained and connecting an output neuron of the NN to input neuron(s) in an input layer of the NN for the selected output; preparing a data set to be learnt by the NN; and, applying the prepared data set to the NN to be learnt by applying an input vector of the prepared data set to the first hidden layer of the NN, or the output layer of the NN if the NN has no hidden layer(s), and determining whether at least one neuron for the selected output in each layer of the NN can learn to produce the associated output for the input vector. If none of the neurons in a layer of the NN can learn to produce the associated output for the input vector, then a new neuron is added to that layer to learn the associated output which could not be learnt by any other neuron in that layer. The new neuron has its output connected to all neurons in next layer that are relevant to the output being trained. If an output neuron cannot learn the input vector, then another neuron is added to the same layer as the current output neuron and all inputs are connected directly to it. This neuron learns the input the old output could not learn. An additional neuron is added to the next layer. The inputs to this neuron are the old output of the NN, and the newly added neuron to that layer.

This invention presents an innovative framework for the application of machine learning for identification of alloys or composites with desired properties of interest. For each output property of interest, we identify the corresponding driving (input) factors. These input factors may include the material composition, heat treatment, process, microstructure, temperature, strain rate, environment or testing mode. Our framework assumes selection of optimization technique suitable for the application at hand and data available, starting with simple linear, or quadratic, regression analysis. We present a physics-based model for predicting the ultimate tensile strength, a model that accounts for physical dependencies, and factors in the underlying physics as a priori information. In case an artificial neural network is deemed suitable, we suggest employing custom kernel functions consistent with the underlying physics, for the purpose of attaining tighter coupling, better prediction, and extracting the most out of the—usually limited—input data available.

A system for detecting whether a file including content s associated with a cyber-attack is described. The content may include an executable file for example. The system includes an intelligence-driven analysis subsystem and a computation analysis subsystem. The intelligence-driven analysis subsystem is configured to (i) receive the file, (ii) inspect and compute features of the file for indicators associated with a cyber-attack, and (iii) produce a first output representing the detected indicators. The computational analysis subsystem includes an artificial neural network to (i) receive a network input being a first representation of at least one section of binary code from the file as input, and (ii) process the first representation of the section to produce a second output. The first output and the second output are used in determination a classification assigned to the file.

The invention provides an electronic nose air quality monitoring system and method based on cloud computing. The system is provided with an electronic nose cloud terminal, a cloud monitoring center and a mobile phone client. A mode recognition network, a concentration computing network, a pollution source positioning network and an air quality forecasting network are configured on the cloud monitoring center. The electronic nose air quality monitoring system has the advantages that system integrates technologies such as an electronic nose, cloud computing, Internet of Things, clean energy and manual neural network, air quality and position information of each monitoring point are collected in real time and stored, information such as pollution gas category, concentration, pollution source position and air quality forecasting of each monitoring point is obtained after algorithm processing, and results are pushed to the mobile phone client, a large-range and large-scale air quality remote real-time monitoring and forecasting function is achieved, the monitoring results are precise, and the monitoring function is powerful.

The invention discloses a predication method for three-dimensional dose distribution in intensity modulated radiation therapy plans. The method includes steps of (1) collecting valid intensity modulated radiation therapy plan data and forming a case database; (2) dividing a PTV and different to-be-endangered organs of patients into a plurality of voxels according to the resolution ratio of CT images; (3) extracting anatomical characteristics of each patient in the database; (4) extracting dose characteristics of each patient in the database; (5) constructing an artificial neural network, inputting the anatomical characteristics and the dose characteristics of the patients, and learning the mapping relation between the anatomical characteristics and the dose characteristics by the aid of the artificial neural network, and obtaining a correlation model of the anatomical characteristics and the dose characteristics; (6) using the correlation model to predicate the three-dimensional dose distribution of a new patient. The application of said method is using the dose distribution predication method for dose prediction for to-be-endangered organs of patients and quality control is achieved. By adopting the above method, predication of three-dimensional dose distribution in intensity modulated radiation therapy plans can be realized and the method can be applied to a quality control link.

The present disclosure provides an artificial neural network system for identifying a lesion in a retinal fundus image that comprises a pre-processing module configured to separately pre-process a target retinal fundus image and a reference retinal fundus image taken from a same person; a first neural network (12) configured to generate a first advanced feature set from the target retinal fundus image; a second neural network (22) configured to generate a second advanced feature set from the reference retinal fundus image; a feature combination module (13) configured to combine the first advanced feature set and the second advanced feature set to form a feature combination set; and a third neural network (14) configured to generate, according to the feature combination set, a diagnosis result. By using a target retinal fundus image and a reference retinal fundus image as independent input information, the artificial neural network may simulate a doctor, determining lesions on the target retinal fundus image using other retinal fundus images from the same person as a reference, thereby enhancing the diagnosis accuracy.

This invention provides a super acoustic online intelligent system for detecting structure defect, which contains super acoustic defect detector comprising synchronous circuit, emission circuit, probe with transducer and receiving circuit, high speed data sampling card connected with synchronous circuit by gate control circuit, and computer, wherein the the logic control circuit connected with gate control circuit and micro processor of computer, the micro processor having each other connected wavelet analysis module and artificial neural network.