Skip to content

 

Effortlessly deploy sub-mW AI-based vision applications

 

We built the CamCube device demonstrator to inspire & show silicon designers that bringing intelligent vision into a power-friendly ambient computing SoC is now possible, but also easy & fast.

camcube blue
Capture-1

CamCube: from IP design platforms to a power-efficient AI-based Vision Device

  • RAPTOR: The Award-winning uLP AI accelerator IP provided with its user-friendly SDK and a growing model zoo

  • CHAMELEON: User-configurable AMBA-based SoC chassis IP platform granting the best energy-efficiency in days instead of months

  • SPIDER: One-stop-shop power management IP platform enabling the safe and fast construct of an energy-efficient power network on SoC

  • BAT: Application-optimized Audio IP platform enabling HiFi and enhanced audio experience lasting for days

  • PANTHER: Scalable AI processor IP bringing audio processing at new heights in terms of power frugality with its user-friendly SDK


 

icon 1

> 20 FPS

High Frame Rate

icon 2

Sub- mW

Ultra-Low Power Consumption

icon 3

> 95 % accuracy

Optimized ML models

icon 4

Quick Integration

Low Code software

icpn 5

Affordable

Low BOM approach

 

CamCube is a small and low-cost hardware module that detects people, tracks hands, performs gesture recognition and enables voice interfaces.

It can operate on battery for days and a standard USB-C port enables powering and integrating it with existing embedded system.

For CES 2023, Dolphin Design & Neovision teamed up to build CamCube, a device-like demonstrator enabling AI-based vision applications at sub-mW level. 

Picture9 (1) (1)