Mobitv, Inc.

Brainworks Foundry, Inc.

Phillip Alvelda, Vii

Todd Allen Stiers

Jeremy S De Bonet

Markus Krause

Zvi Huber

Jeffrey R Annison

Aleksandra Joanna Sokolowska

Paul M Scanlan

The present disclosure relates to systems and methods to implement efficient high-bandwidth shared memory systems particularly suited for parallelizing and operating large scale machine learning and AI computing systems necessary to efficiently process high volume data sets and streams.

Efficient high bandwidth shared memory architectures for parallel machine learning and AI processing of large data sets and streams

Systems and methods for fully-automated contact-less biometric measurement include receiving a stream of image frames from an internet capable device. Frame data is extracted from the stream of image frames. A predictive load balancer selects a worker server based on a load prediction. An interface delivery server communicates the frame data from the internet capable device to the worker server. A set of patient face images captured by the frame data are extracted. A facial recognition machine learning model determines a patient identity associated with the set of patient face images of the frame data. A shared memory system stores the set of patient face images. The worker server determines biometric measurements based on the set of patient face images in the shared memory system using independent biometric data processing pipelines with shared access to the shared memory system for inter-process communication.

Fully automated non-contact remote biometric and health sensing systems, architectures, and methods

A method and system can comprise a software architecture that allows different applications in the same or different communications protocols to interact with shared resources. More specifically, code for a computer program may be written to increase the amount of code that is generic to (i.e., shared by) more than one application or communications protocol and reduce the amount of code that handle application-specific or protocol-specific actions. In one embodiment, a transaction may be broken down into a set of discrete actions. The discrete actions may include functions that are common to more than one application. These functions may be part of the shared resources. Each application may have its own set of software plug-in modules for the discrete actions.

Efficient multi-protocol software architecture with shared resources for different applications

Systems and methods for improving the performance of a data storage and retrieval system by enabling dynamic switching from one internal data structure to another in response to detecting conditions indicating that a switch would improve performance. In one embodiment, a network proxy implements a cache using a first internal data structure. The caches objects comprise Web pages, and the cache keys comprise URLs corresponding to the Web pages. The proxy monitors cache usage and periodically determines costs associated with usage of the first data structure and an alternative data structure. If the costs associated with the alternative data structure are less than the costs associated with the first data structure, the proxy crates the alternative data structure, migrates data from the first data structure to the alternative data structure, begins using the alternative data structure for the cache, and deletes the first data structure.

Growing community of inventors

Oakland, CA, United States of America

Phillip Alvelda, Vii