FoodViz

Inspiration

We live now in a big world with many different languages, cultures, and lifestyles. When it comes to food,every once in a while, everyone gets hungry and wants to grab a bite. We saw that menus in most restaurants have two issues, the items don't have pictures and they are not descriptive enough or the food is from another cuisine, and you don't have a clue of how it looks or what it contains, especially if you are traveling over the world and everything is in a language you don't even know.

What it does

It gives the user the ability to select menu items , in English for now, and have a photo of that item appear on the screen with the option to choose other photos.

How WE built it

The program is built in the ios framework using: Tesseract for character recognition from images CoreML and Vision to identify the bounding boxes of the characters

Challenges WE ran into

Texts that are close together are difficult to select with drag detection, making word detection challenging. Getting a http image search request from Bing had a steep learning curve,so initial progress was frustrating.

Accomplishments that WE'RE proud of

Being able to combine all of the individual parts into a single (mostly) coherent project in the course of 24 hours was a time crunch and physical strain. We are proud that we stayed awake for most of the time and had a presentable project despite time constraints.

What WE learned

Learning Swift in the span of 24 hours was a challenge, but after the projects, every member learned a lot. Learning to make http requests and utilize Teseract, CoreML, and Vision was both difficult and rewarding.

What's next for FoodViz

FoodViz will be able to detect food items from different languages giving the user a well chosen photo of that item, in addition it will provide if that item has any type of food restrictions, such as Gluten-free/ Dairy-free/ Halal / Kosher/ Doesn't contain peanuts, etc ... Moreover, In the future we can create a database that for each restaurant each item in it is photographed and 3d rendered, so the customer can see the item. Because at the end, people expect three things from food: "How it looks, How it smells and How it tastes", and I think we can satisfy the first :)