Isn't it frustrating when you are grocery shopping but don't quite remember if you had that carton of milk in the fridge? How about when your tomatoes go bad when you bought just a couple too many? Do you ever have the trouble when you are just staring at your fridge contents but don't see the ingredients for any decent meal?
Say goodbye to your fridge management problems with FridgeNet! FridgeNet is a sentient AI from the future that will revolutionise how you think about your perishables. We at 6o labs have integrated the latest research in computer vision, internet-of-things and consumer electronics for kitchen delAIght.
- Friendly, Sentient and Superhuman Artificial Fridge Intellingence
- Automatic inventory management integrated with your fridge
- Customised recipe suggestions from top chefs of the world
- Futuristic easy-to-use fridge management interface
Perpetual struggle with figuring out what to eat and what to buy from the grocery store.
What it does
Uses computer vision to identify objects you put in or take out to/from your fridge and updates its inventory. Displays inventory and recipe suggestions.
How I built it
We used google cloud vision API for the object recognition, some other APIs the recipe suggestions. The camera code is written for iphone using swift. The backend uses python and flask.
Challenges I ran into
We tried to build the computer vision algorithm from scratch but turns out training deep neural networks is non-trivial in 48 hours.
Accomplishments that I'm proud of
Integrated and beautiful end-to-end experience.
What I learned
Get a lot of high quality training data for DNNs. Keep it simple stupid. Have fun.
What's next for "YOUR PROJECT"
Getting featured on https://twitter.com/internetofshit