“Maple Syrup Pi Camera is part of my research project,” Ricardo tells us. “I’m one of the 25 TRAIN@Ed Research Fellows at University of Edinburgh. I am working on a project focused on helping local tourist attractions to better manage tourist flow, still GDPR compliant by design. My solution is to process information in the camera without ever saving or transmitting personal data. In addition to that, I wanted something open-source that could be customised, and as low-power as possible, allowing it to run a full day on an off-the-shelf power bank. Raspberry Pi Zero W is my go-to IoT device and, together with the Google Coral USB Accelerator, it has become my powerhouse for machine learning inference.”
The Coral USB Accelerator attached to Raspberry Pi contains an Edge TPU processor which allows for offline TensorFlow models to run. It has some similar tech to the AIY Vision Kit that Google released a little while back.
If you’re interested in reading about how it works in more detail, you can head to Ricardo’s project page.