43 results about "Spatiotemporal pattern" patented technology

A system and method for characterizing a pattern, in which a spiking neural network having at least one layer of neurons is provided. The spiking neural network has a plurality of connected neurons for transmitting signals between the connected neurons. A model for inducing spiking in the neurons is specified. Each neuron is connected to a global regulating unit for transmitting signals between the neuron and the global regulating unit. Each neuron is connected to at least one other neuron for transmitting signals from this neuron to the at least one other neuron, this neuron and the at least one other neuron being on the same layer. Spiking of each neuron is synchronized according to a number of active neurons connected to the neuron. At least one pattern is submitted to the spiking neural network for generating sequences of spikes in the spiking neural network, the sequences of spikes (i) being modulated over time by the synchronization of the spiking and (ii) being regulated by the global regulating unit. The at least one pattern is characterized according to the sequences of spikes generated in the spiking neural network.

Providing enhanced video content includes processing at least one video feed through at least one spatiotemporal pattern recognition algorithm that uses machine learning to develop an understanding of a plurality of events and to determine at least one event type for each of the plurality of events. The event type includes an entry in a relationship library detailing a relationship between two visible features. Extracting and indexing a plurality of video cuts from the video feed is performed based on the at least one event type determined by the understanding that corresponds to an event in the plurality of events detectable in the video cuts. Lastly, automatically and under computer control, an enhanced video content data structure is generated using the extracted plurality of video cuts based on the indexing of the extracted plurality of video cuts.

According to exemplary embodiments of the present disclosure, it is possible to provide method, system, arrangement, computer-accessible medium and device to stimulate individual neurons in brain slices in any arbitrary spatio-temporal pattern, using two-photon uncaging of photo-sensitive compounds such as MNI-glutamate and / or RuBi-Glutamate with beam multiplexing. Such exemplary method and device can have single-cell and three-dimensional precision. For example, by sequentially stimulating up to a thousand potential presynaptic neurons, it is possible to generate detailed functional maps of inputs to a cell. In addition, it is possible to combine this exemplary approach with two-photon calcium imaging in an all-optical method to image and manipulate circuit activity. Further exemplary embodiments of the present disclosure can include a light-weight, compact portable device providing for uses in a wide variety of applications.

A light emitting device, a light receiving device and a data transmission system and a data transmission method using the same are provided. The data transmission system includes the light emitting device and the light receiving device. The light emitting device includes a light emitting unit and a control circuit, and the light receiving device includes an image capture unit, a recognition unit and a decoding unit. The control circuit controls the light emitting unit to sequentially display a plurality of pattern images in a plurality of frame times to form a spatiotemporal pattern image according to setting data corresponding to a temporal domain and a spatial domain. The image capture unit captures the spatiotemporal pattern image. The recognition unit recognizes the spatiotemporal pattern image to output recognized data. The decoding unit decodes the recognized data to output decoded data.

A system for enabling user interaction with video content includes an ingestion facility configured to access at least one video feed and a machine learning system configured to process the at least one video feed through a spatiotemporal pattern recognition algorithm that applies machine learning on an event in the at least one feed in order to develop an understanding of the event including identifying context information relating to the event and an entry in a relationship library at least detailing a relationship between two visible video features. The system further includes an extraction facility configured to automatically extract content displaying the event and associate the extracted content with the context information, and a video production facility configured to produce a video content data structure that includes the context information. The system further includes a user interface configured with video interaction options that are based on the context information.

A spiking neural network has a layer of connected neurons exchanging signals. Each neuron is connected to at least one other neuron. A neuron is active if it spikes at least once during a time interval. Time-varying synaptic weights are computed between each neuron and at least one other neuron connected thereto. These weights are computed according to a number of active neurons that are connected to the neuron. The weights are also computed according to an activity of the spiking neural network during the time interval. Spiking of each neuron is synchronized according to a number of active neurons connected to the neuron and according to the weights. A pattern is submitted to the spiking neural network for generating sequences of spikes, which are modulated over time by the spiking synchronization. The pattern is characterized according to the sequences of spikes generated in the spiking neural network.

Presenting event-specific video content that conforms to a user selection of an event type includes processing at least one video feed through at least one spatiotemporal pattern recognition algorithm that uses machine learning to develop an understanding of at least one event within the at least one video feed to determine at least one event type, wherein the at least one event type includes an entry in a relationship library at least detailing a relationship between two visible features of the at least one video feed, extracting the video content displaying the at least one event and associating the understanding with the video content in a video content data structure. A user interface is configured to permit a user to indicate a preference for at least one event type that is used to retrieve and provide corresponding extracted video content with the data structure in a new video feed.

A space that closely approximates the state of an actual flame is reproduced without depending on temporal periods. Namely, by reproducing a spatiotemporal pattern of a flame, the light source can be caused to emit warm light, whereby a compact and inexpensive imitation flame generating apparatus is provided. The imitation flame generating apparatus 1 comprises a light source 10 and a control device 40 for controlling the output of electric current to the light source 10. The control device 40 comprises computation means 41 for computing a spatiotemporal pattern of the flame using a coupled map lattice, and output means 42 for outputting the electric current in accordance with the thus computed spatiotemporal pattern of the flame.

This invention involves use of temporal or spatio / spector-temporal data (SSTD) for early classification of outputs that are results of spatio-temporal patterns of data. Classification models are based on spiking neural networks (SNN) suitable to learn and classify SSTD. The invention may predict early events in many applications, i.e. engineering, bioinformatics, neuroinformatics, predicting response to treatment of neurological and brain disease, ecology, environment, medicine, and economics, among others. The invention involves a method and system for personalized modelling of SSTD and early prediction of events based on evolving spiking neural network reservoir architecture (eSNNr). The system includes a spike-time encoding module to encode continuous value input information into spike trains, a recurrent 3D SNNr and an eSSN as an output classification module.

A user interface for a media system supports using gestures, such as swiping gestures and taps, to navigate frame-synchronized video clips or video feeds. The detection of the gestures is interpreted as a command to navigate the frame-synchronized content. In one implementation, a tracking system and a trained machine learning system is used to generate the frame synchronized video clips or video feeds. In one implementation, video clips of an event are organized into storylines and the user interface permits navigation between different storylines and within individual storylines.

The invention relates to the technical field of gait recognition, and discloses a gait recognition method based on a dynamic vision sensor. The invention creatively provides a spatiotemporal mode analysis method based on the dynamic vision sensor. Through an impulsive neural network model based on a Tempotron algorithm, the goal of training and recognizing gait data recorded by the dynamic vision sensor can be achieved, so that the finally obtained gait recognition has high biological authenticity; therefore the gait recognition can be performed on a plurality of objects; the problem of high difficultly of the gait detection in the complicated background can be solved; and the high accuracy of the gait recognition can be ensured. Meanwhile, the invention also provides two coding modes; the fast convergence can be realized in the training process; and the good recognition correctness rate is obtained. Particularly, by combining the fixed period mobile window data segment sample division mode, the correctness rate of the gait recognition can reach 85 percent or higher; the practical values are very high; and the practical popularization and application are convenient.

Systems and methods for identifying and analyzing neuropsychological flow patterns, include creating a knowledge base of neuropsychological flow patterns. The knowledge base is formed by obtaining signals from multiple research groups for particular behavioral processes, localizing sources of activity participating in the particular behavioral processes, identifying sets of patterns of brain activity for the behavioral processes and neuropsychologically analyzing the localized sources and the identified patterns for each of the research groups. The neuropsychological analysis includes identifying all possible pathways for the identified sets of patterns, ranking the possible pathways based on likelihood for the particular behavioral process and reducing the number of ranked possible pathways based on additional constraints. A system for comparison of obtained signals from an individual to the created knowledge base is provided. These obtained signals are then used to further update the existing knowledge base.

A system and method for detecting and predicting neurological events with an implantable device uses a relatively low-power central processing unit in connection with signal processing circuitry to identify features (including half waves) and calculate window-based characteristics (including line lengths and areas under the curve of the waveform) in one or more electrographic signals received from a patient's brain. The features and window-based characteristics are employed within the framework of a programmable finite state machine to identify patterns and sequences in and across the electrographic signals, facilitating early and reliable detection and prediction of complex spatiotemporal neurological events in real time, and enabling responsive action by the implantable device.

A video analysis apparatus may include a generating unit to generate at least one spatiotemporal pattern by performing pixel sampling on a plurality of frames in an input video, an extracting unit to extract at least one region of interest having a sinusoidal pattern from the at least one spatiotemporal pattern, and an analysis unit to perform a frequency analysis of the at least one region of interest to determine whether the input video includes a predetermined type of content.

A transaction data analyzer associated with a financial entity discovers patterns and / or sequences in consumer transaction data. The analyzer may provide businesses with feedback on spatiotemporal patterns in consumer spending habits. In certain embodiments, the transaction analyzer discovers the frequency of a sequence of purchases made at a first merchant immediately followed by purchases made at a second merchant. In another embodiment, the transaction analyzer discovers trends in consumer purchases made during the weekday versus those that are made during the weekend. The results of the analysis may be used in a variety of ways, including, but not limited to, risk mitigation, merchant / consumer prospecting, and targeted promotions.

A light emitting device, a light receiving device and a data transmission system and a data transmission method using the same are provided. The data transmission system includes the light emitting device and the light receiving device. The light emitting device includes a light emitting unit and a control circuit, and the light receiving device includes an image capture unit, a recognition unit and a decoding unit. The control circuit controls the light emitting unit to sequentially display a plurality of pattern images in a plurality of frame times to form a spatiotemporal pattern image according to setting data corresponding to a temporal domain and a spatial domain. The image capture unit captures the spatiotemporal pattern image. The recognition unit recognizes the spatiotemporal pattern image to output recognized data. The decoding unit decodes the recognized data to output decoded data.

A system and method for detecting and predicting neurological events with an implantable device uses a relatively low-power central processing unit in connection with signal processing circuitry to identify features (including half waves) and calculate window-based characteristics (including line lengths and areas under the curve of the waveform) in one or more electrographic signals received from a patient's brain. The features and window-based characteristics are employed within the framework of a programmable finite state machine to identify patterns and sequences in and across the electrographic signals, facilitating early and reliable detection and prediction of complex spatiotemporal neurological events in real time, and enabling responsive action by the implantable device.

An exemplary method for analyzing behavior of a system includes receiving dynamical measurement data regarding the system, applying a quadratically constrained quadratic 0-1 problem to identify data among the received data, and storing the identified data. An exemplary seizure warning method includes continuously calculating STLmax values, identifying critical sites by applying a quadratically constrained quadratic 0-1 solution to data (e.g. STLmax profiles) derived from pre-seizure onset and post-seizure onset EEG signals, monitoring a T-index curve of the identified critical sites, warning of an impending seizure, observing a seizure and then repeating the cycle by identifying critical sites using the new data, and then monitoring them.

The invention relates to the technical field of visualization and aims to provide a space-time mode visual analysis system based on air quality data. The system comprises a data preprocessing module,a data analysis module and a visualization module. Visual analysis is mainly explored from time sequence changes of a single city, time sequence changes of different modes of adjacent city groups in geographic space and inter-city space relevance in the modes, and an analysis result is coordinated through multi-view linkage. An interactive view is designed to represent a specific data distributioncharacteristic mode. A rule mode and an abnormal mode are explored in the space-time characteristics. Furthermore, the mode is further analyzed to extract valuable information. An analyst can be helped to visually and comprehensively analyze a conventional mode of air quality data, explore implicit data mode exploration mode distribution characteristics and explore a time-varying trend, decisionsupport is provided for the analyst, and a scientific basis is provided for making an air pollution treatment policy.

Systems and methods for identifying and analyzing neuropsychological flow patterns, include creating a knowledge base of neuropsychological flow patterns. The knowledge base is formed by obtaining signals from multiple research groups for particular behavioral processes, localizing sources of activity participating in the particular behavioral processes, identifying sets of patterns of brain activity for the behavioral processes and neuropsychologically analyzing the localized sources and the identified patterns for each of the research groups. The neuropsychological analysis includes identifying all possible pathways for the identified sets of patterns, ranking the possible pathways based on likelihood for the particular behavioral process and reducing the number of ranked possible pathways based on additional constraints. A system for comparison of obtained signals from an individual to the created knowledge base is provided. These obtained signals are then used to further update the existing knowledge base.

The invention belongs to the field of computer algorithms, and particularly relates to a spatial-temporal pattern mining method based on variable-granularity fast GeoHash encoding. The method comprises the steps that firstly, geographic information is encoded through improved geohash; secondly, a multi-level inquiry tree is constructed; thirdly, large-scale unit time spatial patterns are counted; fourthly, data of spatial-temporal patterns (such as volcanic and black hole patterns) is counted. According to the method, GeoHash encoding is improved, a variable-granularity fast GeoHash encoding scheme is disclosed, GeoHash is a kind of address encoding and can convert a two-dimensional longitude and latitude into a character string which can be used for comparison sorting and comparison, and in practical application, the method is more efficient than direct use of the longitude and latitude.

A method of optimizing runtime spatiotemporal events processing is provided. The method includes the following stages: defining a plurality of spatial contexts or spatiotemporal contexts; specifying a plurality of spatial patterns or spatiotemporal patterns being relationships between events complying with specified policies within the defined spatial contexts and spatiotemporal contexts respectively; identifying events of specified properties out of a stream of incoming events, based on the defined spatial contexts or the defined spatiotemporal contexts; detecting event patterns of specified properties out of the stream of incoming events by grouping the incoming events and applying a pattern matching process based on the specified spatial patterns or spatiotemporal patterns; adaptively generating a spatial evaluation plan that defines an order of processing such that event processing is optimized, based on off-line and runtime data; and processing the incoming events based on the order defined in the spatial evaluation plan.

A system and method includes a waveguide for guiding an energy received from a laser. A resonator couples with the waveguide. A signal generator (pump) generates a spatiotemporal pattern wave in the resonator. The spatiotemporal pattern wave produces a wave interaction. Where an interaction between the signal, the spatiotemporal pattern wave and the wave interaction produce linear non-reciprocal behavior in the signal wave.

A system and a method for directing a camera based on time-series data are disclosed, wherein the time-series data represent atomic activities sensed by sensors in an environment, and wherein each atomic activity includes a time and a location at which the each atomic activity is sensed, comprising: providing a spatio-temporal pattern of the specified atomic activity, wherein the spatio-temporal pattern is based only on the time and the location of the atomic activities, such that a spatio-temporal sequence of the atomic activities forms the specified primitive activity; detecting, in the time-series data, a sensed primitive activity corresponding to the spatio-temporal pattern to produce a result, wherein the detecting is performed by a processor; and directing the camera based on the result.

A structured light pattern including a set of patterns in a sequence is generated by initializing a base pattern. The base pattern includes a sequence of colored stripes such that each subsequence of the colored stripes is unique for a particular size of the subsequence. The base pattern is shifted hierarchically, spatially and temporally a predetermined number of times to generate the set of patterns, wherein each pattern is different spatially and temporally. A unique location of each pixel in a set of images acquired of a scene is determined, while projecting the set of patterns onto the scene, wherein there is one image for each pattern.

The invention provides a method and a system for detecting abnormal events of networks. The method includes learning observed state sequences and implied state sequences in actual deployment of the networks to acquire optimal parameters; combining the optimal parameters to calculate implied detection sequences according to inputted observed detection sequences and detecting the abnormal events of the networks. The observed state sequences and implied state sequences are functions of spatial coordinate values of time slice and sensor nodes. The method and the system have the advantages that the various optimal parameters in random fields under dynamic conditions can be acquired via offline learning procedures and then are combined with one another to detect the abnormal events in an online manner according to the inputted observed detection sequences, so that the abnormal events of the networks within a period of time can be known; the various sequences are temporal and spatial functions, and accordingly the abnormal events of the networks can be temporally and spatially detected in a unified manner by the aid of spatial-temporal mode sets.

Described is a system for learning, prediction, and recall of spatiotemporal patterns. An input spatiotemporal sequence is learned using a recurrent spiking neural network by first processing the input spatiotemporal sequence using the recurrent spiking neural network. The recurrent spiking neural network comprises neurons having excitatory synaptic connections and inhibitory synaptic connections. Balanced inhibitory connectivity exists between neurons having excitatory synaptic connections. The recurrent spiking neural network uses distinct forms of synaptic plasticity for excitatory synaptic connections and inhibitory synaptic connections, such that excitatory synaptic connections strengthen and inhibitory synaptic connections weaken. In another aspect, the system is able to recall the learned spatiotemporal sequence and predict a future spatiotemporal sequence through activation of the recurrent spiking neural network.

A transaction data analyzer associated with a financial entity discovers patterns and / or sequences in consumer transaction data. The analyzer may provide businesses with feedback on spatiotemporal patterns in consumer spending habits. In certain embodiments, the transaction analyzer discovers the frequency of a sequence of purchases made at a first merchant immediately followed by purchases made at a second merchant. In another embodiment, the transaction analyzer discovers trends in consumer purchases made during the weekday versus those that are made during the weekend. The results of the analysis may be used in a variety of ways, including, but not limited to, risk mitigation, merchant / consumer prospecting, and targeted promotions.