Food Detection

• 

  Home Screen

• 

  Predicted output

• 

  File insert

Inspiration

This application is a small step for big project. With the help of food detection we can create a food intake app where this will automatically add into the daily food activity. So that user can have a tack of food what they are intaking each and every day.

What it does

In this web application I have created food detection for three classes i.e., samosa, donut and pizza. due to time stains I couldn't able to train for more class. From my model I'm getting a 99% of accuracy of predicting the food.

How we built it

With the help of Tensorflow I build a model by training three classes( donut, samosa and pizza ) and store the model in h5 file which is used in the frontend for predicting the given images. Here frontend was developed in python flask

Challenges we ran into

Training time was huge for more number of classes

Accomplishments that we're proud of

Was very successful for training for three class with accuracy of 99%

What we learned

Was learned the basic concepts of Deep learning in detecting the various food as well as building a frontend.

What's next for Food Detection

Create a mobile app where we add the tracking system of what they eat every day with just in a ease of click the photo.

Built With

• deep-learning
• flask
• keras
• python
• tensorflow