I definitely have not tried this myself but that would be very interesting. It should definitely be possible to implement the device-side of VAS on Arduino. This is an Apple Pay example but might help: https://stackoverflow.com/questions/28497546/which-aid-does-apple-pay-respond-to
Unfortunately I don’t have access to share my dataset. It was also highly specific just to the beer I was training on. You could take my approach and take a lot of photos, or alternatively you could write a scraper or similar to scrape a structured source of photos (like Google Images).
Thanks David! Google’s Cloud Vision API is pre-trained and (so far as I am aware) cannot be retrained with new data. The Vision API does a really good job, then, of recognising a beer bottle, but it can’t tell me what brand of beer it is looking at. With our home-trained model we were able to provide our own training data to allow the algorithm to…
We recently helped Coles in Australia build a great list experience leveraging the hiku device, though there are still leaps to be made. In Accenture Technology Labs we’ve been thinking a lot about how we can personalise the shopping experience in many of the ways described here.