Inspiration

Grocery shopping with my father, we realized we couldn't find thing right under our noses!

How it works

Right now, it runs a little server that waits for an image, which is then chopped into rows of items (shelves) and sent to www.metamind.io to see if they contain a target food item, which is sadly not configurable right now.

Challenges I ran into

Training the MetaMind AI requires a lot of data!

Accomplishments that I'm proud of

We were able to correctly identify Doritos and Cheerios on certain shelves, and identify when they were not there, most of the time.

What I learned

Machine Learning requires a lot of data!

What's next for Food Finder

Accuracy improvements, flexibility, and an Android app.

Share this project:

Updates