camera recognition and AdHawk Mindlink glasses controls for our project, SightShop
by Andy, Derek, Derrick, Owen
Prices for goods across the country have been rapidly increasing over the past couple of years putting a strain on people's wallets. We wanted to help alleviate the pressure that so many Canadians have by building an application that listed multiple prices for the same good allowing for a more informed decision.
SightShop uses the built in camera feature of Adhawk's glasses to identify objects within a user's point of view. SightShop then surfs the internet for identified object and lists them out on a table in a separate page for users to compare. The table includes prices, store, and product name.
Adhawk Mindlink tracking data for user controls (blink event) For object identification, we used YOLOv3 pre-trained model to identify objects with the Mindlink cameras. This is then relayed onto a flask web app where the table will also be included on a separate page.
One of the biggest challenges we faced was the merging of the camera live stream and the web app itself. There was a lot of trouble setting up the Mindlink itself
We're proud of being able to produce a solution using new hardware presented at the booth (Adhawk Mindlink).
We learnt to manipulate Flask and object identification models.
A next step for SightShop is to utilize pupil movement software within the Adhawk Mindlink glasses for a more accurate reading of user's intent for greater accuracy and convenience.