A quick run through of the recently released Coral Accelerator running a demo on a Raspberry Pi with a Pi Camera, using the Edge TPU to recognise household items.
This means you can do machine learning tasks locally without sending data off to the cloud (enabling offline usage, faster responses and working with private data)
From this initial look it's clearly fast, efficient and has lots of potential!
Details for the Accelerator itself and the demo are available on the Coral website here: https://coral.withgoogle.com and a similar tutorial (with some extra electronics) is on the MagPi site here: https://www.raspberrypi.org/magpi/teachable-machine-coral-usb-accelerator/
(and ignore my mistake near the end about the MagPi - it IS the official RPi magazine after all
