Coral USB Accelerator

The Coral USB accelerator brings machine learning inferencing to existing systems. Works with Raspberry Pi and other Linux systems.

Features

• Performs high-speed ML inferencing: High-speed TensorFlow Lite inferencing with low power, small footprint, local inferencing

• Supports all major platforms: Connects via USB 3.0 Type-C to any system running Debian Linux (including Raspberry Pi), macOS, or Windows 10

• Supports TensorFlow Lite: no need to build models from the ground up. Tensorflow Lite models can be compiled to run on the edge TPE

• Supports AutoML Vision Edge: easily build and deploy fast, high-accuracy custom image classification models at the edge. 

• Compatible with Google Cloud

The on-board Edge TPU is a small ASIC designed by Google that accelerates TensorFlow Lite models in a power-efficient manner: it's capable of performing 4 trillion operations per second (4 TOPS), using 2 watts of power—that's 2 TOPS per watt. For example, one Edge TPU can execute state-of-the-art mobile vision models such as MobileNet v2 at almost 400 frames per second. This on-device ML processing reduces latency, increases data privacy, and removes the need for a constant internet connection.

This allows you to add fast ML inferencing to your embedded AI devices in a power-efficient and privacy-preserving way. Models can be developed in TensorFlow Lite and then compiled to run on the USB Accelerator.

Application

AI-enabled NVR system

If you are planning to use Coral USB Accelerator for Home Assistant of home automation applications, we recommend Odyssey Blue, an Intel Celeron J4125 powered X86 Windows/Linux mini PC, you can set them together with ip cameras for a local AI processed NVR system. 

Frigate is a completely open source and local NVR designed for Home Assistant with AI-powered object detection. It uses OpenCV and Tensorflow to perform real-time object detection locally for IP cameras. It brings a rich set of features including video recording, re-streaming, and motion detection, and supports multiprocessing. 

System requirements 

• A computer with one of the following operating systems:
  • Linux Debian 10, or a derivative thereof (such as Ubuntu 18.04), and system architecture of either x86-64, Armv7 (32-bit), or Armv8 (64-bit) (Raspberry Pi is supported, but Seeed have only tested Raspberry Pi 3 Model B+ and Raspberry Pi 4)
  • macOS 10.15, with either MacPorts or Homebrew installed
  • Windows 10
• One available USB port (for the best performance, use a USB 3.0 port)
• Python 3.5, 3.6, or 3.7

Resources

• Datasheet
• Get started guide