Hrl Laboratories, LLC

The Boeing Company

Nigel D Stepp

Tsai-Ching Lu

Aruna Jammalamadaka

Praveen K Pilly

Jose Cruz-Albrecht

Vincent De Sapio

Charles E Martin

Qin Jiang

David J Huber

Steven W Skorheim

Heiko Hoffmann

Soheil Kolouri

Franz David Betz

Youngkwan Cho

Ruggero Scorcioni

Deepak Khosla

Kyungnam Kim

Michael D Howard

Kang-Yu Ni

Samuel D Johnson

Nicholas A Ketz

Sean Soleyman

Hyun (Tiffany) J Kim

John Richard Haley, Jr

Richard J Patrick

Braden Mailloux

Thomas M Trostel

Hao-Yuan Chang

Hyun J Kim

Described is a system and method for robotic perception and action. The system includes a mobile autonomous platform programmed to receive a visual input related to an environment and perform a movement sequence that affects the environment based on the visual input. A neural network is in communication with the mobile autonomous platform. The neural network includes a vision autoencoder trained, based on movements of the mobile autonomous platform, to encode visual features of an object in the visual input onto a visual latent space. In addition, the neural network includes an action autoencoder trained, based on movements of the mobile autonomous platform, to encode action features of the movement sequence onto an action latent space. The neural network also includes a mapping layer between the vision autoencoder and the action autoencoder. The mapping layer decodes the visual inputs into an action output of the mobile autonomous platform.

Automatic learning of relevant actions in mobile robots

A method includes obtaining a plurality of data sets, where each data set of the plurality of data sets includes multivariate time series data for a respective sample period of a plurality of sample periods. The method also includes, for each data set of the plurality of data sets, determining recurrence data indicative of recurrent states in the data set and determining, based on the recurrence data, determinism values of a determinism metric and laminarity values of a laminarity metric. The method further includes determining that a particular data set of the plurality of data sets includes data representing an anomalous state based on a determinism-laminarity curve representing the particular data set, where the determinism-laminarity curve is based on the determinism values of the particular data set and the laminarity values of the particular data set. The method also includes generating output data indicating the anomalous state.

Detection of anomalous states in multivariate data

A method for generating human-machine hybrid predictions of answers to forecasting problems includes: parsing text of an individual forecasting problem to identify keywords; generating machine models based on the keywords; scraping data sources based on the keywords to collect scraped data relevant to the individual forecasting problem; providing the scraped data to the machine models; receiving machine predictions of answers to the individual forecasting problem from the machine models based on the scraped data; providing, by the computer system via a user interface, the scraped data to human participants; receiving, by the computer system via the user interface, human predictions of answers to the individual forecasting problem from the human participants; aggregating the machine predictions with the human predictions to generate aggregated predictions; and generating and outputting a hybrid prediction based on the aggregated predictions.

System and method for human-machine hybrid prediction of events

Described is a system for specifying control of a device based on a Bayesian network model. The system includes a Bayesian neuromorphic compiler having a network composition module having probabilistic computation units (PCUs) arranged in a hierarchical composition containing multi-level dependencies. The Bayesian neuromorphic compiler receives a Bayesian network model as input and produces a spiking neural network topology and configuration that implements the Bayesian network model. The network composition module learns conditional probabilities of the Bayesian network model. The system computes a conditional probability and controls a device based on the computed conditional probability.

Network composition module for a bayesian neuromorphic compiler

A method includes obtaining multiple sets of trajectory data, each descriptive of trajectories of two or more objects (e.g., first and second objects). The method also includes generating transformed trajectory data based on the trajectory data. Each set of transformed trajectory data is descriptive of the trajectories of the two or more objects in a normalized reference frame in which a movement path of the first object is constrained. The method further includes generating feature data, performing a clustering operation based on the feature data to generate a set of trajectory clusters, and generating training data based on the set of trajectory clusters. The method further includes using the training data to train a machine learning classifier to classify particular trajectory patterns.

Trajectory classification and response

Described is a system for computing conditional probabilities of random variables for Bayesian inference. The system implements a spiking neural network of neurons to compute the conditional probability of two random variables X and Y. The spiking neural network includes an increment path for a synaptic weight that is proportional to a product of the synaptic weight and a probability of X, a decrement path for the synaptic weight that is proportional to a probability of X, Y, and delay and spike timing dependent plasticity (STDP) parameters such that the synaptic weight increases and decreases with the same magnitude for a single firing event.

Spiking neural network for probabilistic computation

A method of training an artificial neural network having a series of layers and at least one weight matrix encoding connection weights between neurons in successive layers. The method includes receiving, at an input layer of the series of layers, at least one input, generating, at an output layer of the series of layers, at least one output based on the at least one input, generating a reward based on a comparison of between the at least one output and a desired output, and modifying the connection weights based on the reward. Modifying the connection weights includes maintaining a sum of synaptic input weights to each neuron to be substantially constant and maintaining a sum of synaptic output weights from each neuron to be substantially constant.

Artificial neural networks having competitive reward modulated spike time dependent plasticity and methods of training the same

Described is a system for performing probabilistic computations on mobile platform sensor data. The system translates a Bayesian model representing input mobile platform sensor data to a spiking neuronal network unit that implements the Bayesian model. Using the spiking neuronal network unit, conditional probabilities are computed for the input mobile platform sensor data, where the input mobile platform sensor data is a time series of mobile platform error codes encoded as neuronal spikes. The neuronal spikes are decoded and represent a mobile platform failure mode. The system causes the mobile platform to initiate a mitigation action based on the mobile platform failure mode.

Programming model for a bayesian neuromorphic compiler

A method for collecting and processing user input. In some embodiments the method includes presenting a first user with a prompt for eliciting a first response, the first response including a numerical portion including one or more numbers, and an explanatory portion; receiving, from the first user, the first response; receiving from each of a plurality of other users, a respective response of a plurality of other responses; and displaying, to the first user, an ordered list of other responses. Within the ordered list, a second response, of the plurality of other responses, may be earlier than a third response, of the plurality of other responses, the second response being, according to a measure of distance, more distant, than the third response, from the first response.

System of structured argumentation for asynchronous collaboration and machine-based arbitration

Described is a system for online vehicle recognition in an autonomous driving environment. Using a learning network comprising an unsupervised learning component and a supervised learning component, images of moving vehicles extracted from videos captured in the autonomous driving environment are learned and classified. Vehicle feature data is extracted from input moving vehicle images. The extracted vehicle feature data is clustered into different vehicle classes using the unsupervised learning component. Vehicle class labels for the different vehicle classes are generated using the supervised learning component. Based on a vehicle class label for a moving vehicle in the autonomous driving environment, the system selects an action to be performed by the autonomous vehicle, and causes the selected action to be performed by the autonomous vehicle in the autonomous driving environment.

Method of real time vehicle recognition with neuromorphic computing network for autonomous driving

A system configured to identify a target in a synthetic aperture radar signal includes: a feature extractor configured to extract a plurality of features from the synthetic aperture radar signal; an input spiking neural network configured to encode the features as a first plurality of spiking signals; a multi-layer recurrent neural network configured to compute a second plurality of spiking signals based on the first plurality of spiking signals; a readout neural layer configured to compute a signal identifier based on the second plurality of spiking signals; and an output configured to output the signal identifier, the signal identifier identifying the target.

System and method for synthetic aperture radar target recognition using multi-layer, recurrent spiking neuromorphic networks

A method for retraining an artificial neural network trained on data from an old task includes training the artificial neural network on data from a new task different than the old task, calculating a drift, utilizing Sliced Wasserstein Distance, in activation distributions of a series of hidden layer nodes during the training of the artificial neural network with the new task, calculating a number of additional nodes to add to at least one hidden layer based on the drift in the activation distributions, resetting connection weights between input layer nodes, hidden layer nodes, and output layer nodes to values before the training of the artificial neural network on the data from the new task, adding the additional nodes to the at least one hidden layer, and training the artificial neural network on data from the new task.

Artificial neural network and method of training an artificial neural network with epigenetic neurogenesis

Described is a low power system for mobile devices that provides continuous, behavior-based security validation of mobile device applications using neuromorphic hardware. A mobile device comprises a neuromorphic hardware component that runs on the mobile device for continuously monitoring time series related to individual mobile device application behaviors, detecting and classifying pattern anomalies associated with a known malware threat in the time series related to individual mobile device application behaviors, and generating an alert related to the known malware threat. The mobile device identifies pattern anomalies in dependency relationships of mobile device inter-application and intra-applications communications, detects pattern anomalies associated with new malware threats, and isolates a mobile device application having a risk of malware above a predetermined threshold relative to a risk management policy.

System for continuous validation and threat protection of mobile applications

A system configured to identify a target in a synthetic aperture radar signal includes: a feature extractor configured to extract a plurality of features from the synthetic aperture radar signal; a spiking neural network configured to encode the features as a plurality of spiking signals; a readout neural layer configured to compute a signal identifier based on the spiking signals; and an output configured to output the signal identifier, the signal identifier identifying the target.

System and method for synthetic aperture radar target recognition utilizing spiking neuromorphic networks

Described is neuromorphic system for authorized user detection. The system includes a client device comprising a plurality of sensor types providing streaming sensor data and one or more processors. The one or more processors include an input processing component and an output processing component. A neuromorphic electronic component is embedded in or on the client device for continuously monitoring the streaming sensor data and generating out-spikes based on the streaming sensor data. Further, the output processing component classifies the streaming sensor data based on the out-spikes to detect an anomalous signal and classify the anomalous signal.

Growing community of inventors

Santa Monica, CA, United States of America

Nigel D Stepp