Google launched their maker friendly AIY projects with a Voice Kit back in May, today they’ve launched a new kit with the AIY Vision Kit adding a visual aspect to the project.
Google says the AIY Vision Kit features ‘on-device neural network acceleration, providing powerful computer vision without a cloud connection’. The AIY Vision Kit is again powered by a Raspberry Pi, though this time it’s the Raspberry Pi Zero W board, with a Raspberry Pi Camera and an SD card also required in addition to the AIY Vision Kit.
The AIY Vision Kit includes a number of other components including the VisionBonnet board for Raspberry Pi. The VisionBonnet board has an Intel Movidius MA2450 low-power vision processing unit that Google says is capable of running neural network models on-device that allows the project to work without an internet connection.
Google has provided a software package for the project:
The provided software includes three TensorFlow-based neural network models for different vision applications. One based on MobileNets can recognize a thousand common objects, a second can recognize faces and their expressions and the third is a person, cat and dog detector. We’ve also included a tool to compile models for Vision Kit, so you can train and retrain models with TensorFlow on your workstation or any cloud service.
Google says the completed project will be able to:
- Identify all kinds of plant and animal species
- See when your dog is at the back door
- See when your car left the driveway
- See that your guests are delighted by your holiday decorations
- See when your little brother comes into your room (sound the alarm!)
Like their AIY Voice Kit, Google has again partnered with US-based MicroCenter to distribute the AIY Vision Kit, which will cost $44.99USD and be available in-store from early December. No international suppliers have been announced, nor has any partnership with MagPi magazine been announced, so it looks like this will remain a US-only project for now.