Inspiration
We noticed people, including some of our team members, having difficulty in quickly identifying certain ingredients that they cannot eat that are listed on food labels. We wanted to make a program that would allow these people rapidly and efficiently figure out what items and they can and cannot eat.
What it does
Runs as a web app with a camera that allows users to hold up a food label in front of it. This is linked to a self-created AI that has been trained on food labels that should recognise all allergens that the user has listed to avoid. They are able to list what they have to avoid through an interactive form on the web app. The web app also contains a way for users to permanently add their restrictions by creating an account using a .txt file though down the line we wish to change to use a SQL database when we have more time.
How we built it
We split the tasks in three different categories: the AI, the back-end of the web interface and the front-end of the web interface.
Challenges we ran into
We made a SQL database successfully to store user information however when it come to hosting the database on a web server our progress has halted since either the software crashed and failed to run properly or it required money which we were not willing to pay.
Another challenge we ran into was the implementation of a camera in a web app. As the aim of using a web app over an application for certain operating systems was that anyone with any device can use it, the website had to be written so that the JavaScript senses the camera of the device it is running on. This was a massive challenge as none of us had ever created something like this before, but after some experimentation and perseverance we managed to get it working.
Accomplishments that we're proud of
We are proud to have made a full image recognition AI from scratch as well as an interactive user interface on a web page. Also, we are happy to have made the code for the SQL database though the system was not implemented in the end.
What we learned
All of us on the team have all learnt new programming skills as well as gained new or more experience in at least 1 language.
What's next for J'Eviter
This software could be expanded to various different industries such as healthcare and reading medicine labels. We can see our AI training on more data sets and expanding into being used on various different types of embedded systems.
Log in or sign up for Devpost to join the conversation.