Verizon Patent and Licensing Inc.

Csaba Rekeczky

Istvan Horvath

Domonkos Huszar

Andrew W Herson

Christopher Kennedy

Akos Kiss

Andras Horvath

Felician Benda

Dora E Babicz

Gellert Nacsa

Gabor Majoros

Peter Ruzsa

Anna E Csorgo

Balint Magyar

Gabor Erdosi

Mark Szabo

Techniques described herein provide for the use of a heterogeneous artificial intelligence/machine learning ('AI/ML') architecture, in which relatively complex AI/ML, techniques may be used in conjunction with more lightweight AI/ML techniques in order to leverage the accuracy of relatively complex AI/ML techniques with the reduced processing power and/or time requirements of more lightweight AI/ML techniques. A teacher model system may utilize processing resource and/or time-intensive AI/ML techniques and/or models in order to determine classifications associated with source data, and may provide such classifications to a student model system that may utilize the classifications in accordance with less processing resource and/or time-intensive AI/ML techniques in order to accurately classify sensor data in real time or near-real time.

Systems and methods for domain-specific enhancement of real-time models through edge-based learning

A device may receive a video and corresponding camera information associated with a camera that captured the video, and may select an object in the video and a wire model for the object. The device may adjust an orientation, location, or size of the wire model to align the wire model on the object in a frame of the video, based on the corresponding camera information and to generate an adjusted wire model. The device may identify the object in another frame of the video, and may align the adjusted wire model on the object in the other frame. The device may interpolate the adjusted wire model for the object for intermediate frames of the video between the first and other frames, and may generate three-dimensional annotations for the video based on the adjusted wire models. The device may train a machine learning model based on the three-dimensional annotations.

Systems and methods for generating three-dimensional annotations for training a machine learning model

A method, a device, and a non-transitory storage medium are described in which a machine learning-based device placement and configuration service is provided. The machine learning-based device placement and configuration system uses regression models calculated from installation and performance data. The regression model includes data associated with sites at which video cameras were previously installed and tested for detection accuracy levels possibly associated with a service area. With these generated models, information about the physical space, and desired performance criteria, designers optimize camera number, camera placement, and geo-location camera parameters. The device placement and configuration service may find an optimal camera placement and geo-location camera parameter set which satisfy certain criteria. The geo-location parameters include position, height, heading, pitch, and roll. Subsequent to the installation of the video cameras using the calculated geo-location parameters, the system may verify accuracy of detection of the service area, and update the regression model.

Machine learning-based device placement and configuration service

A device may receive initial registration data identifying initial camera parameters associated with a camera device provided at a location, and may receive location data associated with the location captured by the camera device. The device may receive map data identifying a map image of the location, and may transform the initial registration data into estimated camera parameters. The device may process the location data, with the estimator model, to generate extracted data, and may process the estimated camera parameters, the extracted data, and the map data, with a parameter optimization model, to identify camera parameters for the camera device. The device may provide the camera parameters to the camera device to cause the camera device to be configured based on the camera parameters.

Systems and methods for utilizing modeling to automatically determine configuration parameters for cameras

Growing community of inventors

Monte Sereno, CA, United States of America

Csaba Rekeczky