Figuring out what to cook quickly for today’s dinner is one of the terrifying questions we face every day. With countless hours spent figuring out what to cook today that could have spent instead of binge-watching Netflix or figuring out how awesome you can become.
So to solve this problem for once we set out on the quest to develop a Mobile phone App which is filled with Awesomeness of Augmented reality and Artificial intelligence.
how does our App will help you may ask?
So this is how your journey will look like :
You are out there shopping in a Grocery shop nearby, You browse and crawl endless shelves mindlessly with no idea of what to cook tonight, You pick one product, You know it looks interesting but these are typical questions you have in your mind:
- What recipes can I cook using this product that will take less than 30 minutes to cook?
- How can I cook those recipes?
- What else I need to buy apart from this product?
- Can I share recipes with my Mom and get her advice?
We set out to develop an app that can answer all of these questions. A typical journey of the system starts with a person opening the app, At homepage of the app, you are presented with two options, AR Scan Mode and Show all recipes. The most interesting option of our app is AR Scan mode. In this mode you start browsing store aisle with AR mode, Our unique object detection system ( Based on Deep neural networks) detects items on a shelf and gives visual cues in form of chef’s hat that you can interact with. On clicking these interactive AR pop-ups you are taken to screen where you can see all interesting recipes you can cook using that item. You can then scan and pick missing ingredients from shelf to get your awesome dinner. In the show all recipes mode you are shown the most popular recipes you can cook based on your tastes and preferences. We have set default time-to-cook to 30 Minutes, So by default, it will only show recipes that can be cooked in 30 minutes and less. (You can customize this in preferences).You just pick one and start shopping for missing ingredients from Store. App also has social sharing enabled so you can ask your relatives for advice and feedback.
Technology wise this prototype uses iPhone AR, core M, and vision features to learn new recipes for Dinner. Deep learning model on mobile is bleeding edge of the technology. All of the inference happens on the device itself so no personal data like camera images leaves the phone.
We believe that actions speak louder than words hence instead of theorizing we have made a prototype of our idea. Here is a screen recording of the app in action.