A lot of times we just grab food without keeping a track of what we are consuming or how many calories we intake. College students often have high stress lifestyles, we often lose track of whether we have been eating too much or too little. Grabbing snacks one after another while brainstorming for ideas, we thought of just tracking the calories we have had consumed so far. Our Goal was to make ourselves and others more aware about eating habits so that we can adjust to healthier diets.

What it does

Healthy Hack tries to make it as easy as possible to keep track of what we are eating on a daily basis. You take a snap of your food and Healthy-hack recognizes the picture using deep learning. Further, the calories of the snack is been added to the user's daily calorie count value. Back end, deep learning; everything was made in a Web Service API. Once the calories are added to the user's daily value it is then transferred to Amazon AWS S3 Database. This database is further accessed by Alexa skill set where one can ask the Amazon Echo device, various details about their calorie consumption.

How we built it

Front end website --> Web service API : Python Flask (Clicks and sends Image) --> Tensor-flow Image Classification --> Store/Add Json using API ---> Upload to S3 --> Return Response to User

Alexa--> Alexa Skills-->Lambda Function --> AmazonS3--> User Output

Challenges we ran into

Finding/Generating the correct deep learning data-set for Image Classifications.
Integrating the deep learning model into the Web Service API.
Figuring out the working of AWS was confusing.
Connecting front end to back end Flask API

Accomplishments that we're proud of

Implementation of novel architecture idea. Our team had a discussion for about five hours for coming up with the project idea and its implementation. This made our later process very smooth. We have incorporated multiple technologies, many of which were completely new to all of us, so as to create a complete and cohesive application.

What we learned

Alexa Skills.
Incorporating the Deep Learning model in Web-Service.
Team Integrity.
Mid-Night coding.
Hackathons are fun.

What's next for Healthy-Hack.

We'll probably expand our data set and crowd source more information about more food items.
There is a possibility to create user accounts and safely store individual user data.
We also plan to create a mobile app for the same.

Built With

Share this project: