Cisco Technology, Inc.

Other

Qrs Music Technologies, Inc.

Joseph T Friel

J William Mauchly

Dihong Tian

Hugo Latapie

Enzo Fenoglio

Andre Surcouf

Wen-hsiung Chen

Maurice J Buttimer

Pascal Thubert

Pete Rai

Andrew P Weir

Andre Jean-Marie Surcouf

Michael A Arnao

Alan J MacDonald

Plamen Nedeltchev

Richard Thayer Wales

David J Mackie

Manikandan Kesavan

Wei Ey Li

Toerless Eckert

Philip Roland Graham

Altan John Stalker

Tor A Sundsbarm

Jinshi Huang

Thierry Gruszka

Michael B Costello

David W Sowa

Guido Van Den Berghe

Indrajit Rajeev Gajendran

Olin D Williford, Ii

Pawel Paszkowski

William Mauchly

Wen-Hsiung Chen

Paul J Randall

William J Mauchly

A method comprises collecting, by a computing device located at an edge of a network, data items corresponding to information transmitted by endpoints using the network, generating, by the computing device, a probabilistic hierarchy using the data items, generating, by the computing device using the probabilistic hierarchy and natural language data, a similarity metric, generating, by the computing device using the probabilistic hierarchy, the natural language data, and the similarity metric, an ontology, detecting, by the computing device using the ontology, an anomaly, and in response to detecting the anomaly, sending a notification.

Automatic ontology generation for internet of things applications

In one embodiment, a device in a network monitors a plurality of traffic flows in the network. The device extracts a plurality of features from the monitored plurality of traffic flows. The device generates a context model by using deep learning and reinforcement learning on the plurality of features extracted from the monitored traffic flows. The device applies the context model to a particular traffic flow associated with a client, to determine a context for the particular traffic flow. The device personalizes data sent to the client from a remote source based on the determined context.

Contextual services in a network using a deep learning agent

This disclosure describes a method and system to perform object detection and 6D pose estimation. The system comprises a database of 3D models, a CNN-based object detector, multiview pose verification, and a hard example generator for CNN training. The accuracy of that detection and estimation can be iteratively improved by retraining the CNN with increasingly hard ground truth examples. The additional images are detected and annotated by an automatic process of pose estimation and verification.

Multiview estimation of 6D pose

Various implementations disclosed herein provide a search engine that receives a search request from a sensor gateway, and provides search results in return. In various implementations, the search request includes a first set of measurements captured by a first sensor, a first measurement from the first set of measurements is outside a defined range. In various implementations, the search engine determines a first feature vector based on the first set of measurements, and identifies a second feature vector that indicates a second set of measurements within a degree of similarity to the first set of measurements. In some implementations, the second set of measurements are captured by a second sensor. In various implementations, the search engine determines a search result based on the second feature vector, and transmits the search result. In some implementations, the search result indicates one or more instructions executable by the first sensor.

Search engine for sensors

In one embodiment, a method includes training a deep neural network using a first set of network characteristics corresponding to a first time and a second set of network characteristics corresponding to a second time, generating, using the deep neural network, a predictive set of network characteristics corresponding to a future time, and assigning a task of a distributed application to a processing unit based on the predictive set of network characteristics.

Deep learning application distribution

In one embodiment, a method includes receiving current data, the current data including time series data representing a plurality of time instances. The method includes storing at least a recent portion of the current data in a buffer. The method includes reducing the dimensionality of the current data to generate dimensionality-reduced data. The method includes generating a reconstruction error based on the dimensionality-reduced data and a plurality of neural network metrics. At least one of a size of the recent portion of the current data stored in the buffer or an amount of the reducing the dimensionality of the current data is based on the reconstruction error.

Scalable deep learning video analytics

A system for collision avoidance includes memory storing instructions which, when executed, cause one or more processors to perform determining a direction of flight of a first drone, causing broadcasting, in the direction of flight based, a beamformed signal of beacon frames, determining a new flight direction of the same first drone, in response to the new flight direction, causing broadcasting of the beacon frames in the new flight direction, detecting second beacon frames from a second drone associated with a direction from which the second beacon frames are arriving; in response, causing the first drone to perform, without input from a pilot, one or more of a change in elevation, heading, speed, or type of operation, directed toward causing the first drone to follow a flight path that is separated from the second drone.

Unmanned aerial vehicle collision avoidance system

According to one aspect, a method includes identifying at least a first chunk to be obtained, the at least first chunk including at least a first packet, and determining a deadline for the first chunk, the deadline being indicative of an amount of time before the first chunk is needed. The method also includes determining whether the deadline for the first chunk is relatively long, and de-prioritizing the first chunk with respect to obtaining the first chunk for queueing in a buffer when it is determined that the deadline for the first chunk is relatively long. Finally, the method includes obtaining the first chunk for queueing in the buffer, wherein obtaining the first chunk for queueing in the buffer includes obtaining the first chunk after obtaining a second chunk for queueing in the buffer, the second chunk having a shorter deadline than the deadline for the first chunk.

Method and apparatus for providing deadline-based segmentation for video traffic

In one embodiment, a method includes receiving training data, the training data including training video data representing video of a location in a quiescent state, training a neural network using the training data to obtain a plurality of metrics, receiving current data, the current data including current video data representing video of the location at a current time period, generating a reconstruction error based on the plurality of metrics and the current video data in the embedded space, and generating, in response to determining that the reconstruction error is greater than a threshold, a notification indicative of the location being in a non-quiescent state.

Various implementations disclosed herein provide a method for performing one or more transactions between application containers. In various implementations, the method includes transmitting a key request to a first network node within a cluster of network nodes that are configured to generate and maintain a distributed ledger. In some implementations, the key request indicates that the requested key is for one or more transactions between a first application container and a second application container. In various implementations, the method includes receiving a key in response to transmitting the key request. In some implementations, the key is valid for the one or more transactions between the first application container and the second application container. In various implementations, the method includes synthesizing, at the first application container, transaction data with the key. In various implementations, the method includes transmitting, by the first application container, the transaction data to the second application container.

Performing transactions between application containers

Disclosed are systems, methods, and computer-readable storage media for adaptive telemetry based on in-network cross domain intelligence. A telemetry server can receive at least a first telemetry data stream and a second telemetry data stream. The first telemetry data stream can provide data collected from a first data source and the second telemetry data stream can provide data collected from a second data source. The telemetry server can determine correlations between the first telemetry data stream and the second telemetry data stream that indicate redundancies between data included in the first telemetry data stream and the second telemetry data stream, and then adjust, based on the correlations between the first telemetry data stream and the second telemetry data stream, data collection of the second telemetry data stream to reduce redundant data included in the first telemetry data stream and the second telemetry data stream.

Adaptive telemetry based on in-network cross domain intelligence

In some implementations a method includes receiving a first message from a computing device via a first network. The first message may indicate that the computing device is unable to communicate with a second network. The method also includes determining whether a beacon operator has requested tracking of the computing device. The method further includes transmitting a second message to the beacon operator when the beacon operator has requested tracking of the computing device. The second message may include a first geographical location identified by the first network.

Methods and apparatus for determining the locations of computing devices

Growing community of inventors

Ardmore, PA, United States of America

Joseph T Friel