Imagine you go to a restaurant with your friends. Assume you are a fitness freak and though you have come to the party to enjoy the buffet, as a fitness freak, you are calorie conscious and don't want to go overboard.

Now, imagine you have a mobile application that comes to your rescue: it takes a picture of the dish, identifies its ingredients, and calculates the caloric value of the food! You could take a picture of every dish and calculate its caloric value and can then decide whether to put it on your plate. Further, this app keeps on learning the different dishes that you take pictures of and continues to learn and master itself in this trade so that it can take very good care of your health.

I can see the sparkle in your eyes. Yes, this is the mobile application we want to try in this chapter. We also want to utilize both TensorFlow and Core ML to accomplish this activity. We will be performing the following steps to create the application that we just discussed:

1. Create the TensorFlow image recognition model
2. Convert it into a .ml model file
3. Create an iOS/SWIFT app to use that model

We will go through each of these steps in detail in the upcoming sections.