Deep Vision

Pioneered in large data centers, Deep Learning-powered computer vision is now being deployed across a variety of embedded platforms such as drones, robots, IoT smart cameras and cars. These devices require computer vision capabilities that run at the edge, in real-time and enable long battery life. Deep Vision is meeting these needs with a completely new approach to processors and software.

Our unique vision-specific processor is faster and more power efficient than GPU and DSP processing for modern computer vision problems. Coupled with a full stack of low power and real time deep learning computer vision software, our system gives small devices like drones, robots, and IOT cameras the ability to have embedded computer vision.

Deep Vision has an ultra low power processor that is built from the ground up to run Deep Learning and other computer vision algorithms. Offering 100% programability, our processor can support all computer vision CNNs. We’ve built a full stack of complementary software that enable low power and real time processing.

Innovative Architecture
*High performance, scalable processor architecture that is completely programmable
*60x better performance vs. competitors.

Deep Learning Support
*Built specifically to support Deep Learning, CNNs and traditional computer vision algorithms.
*Software for easy software design and implementation.

Low Power Design
*Hardware and software combine to provide unprecedented power savings without performance loss.
*Ideal for both embedded power and larger scalable solutions.