Considering the above mentioned challenges, IBM Watson Studio helps us to get started anyway by offering (right out of the box) the Watson Visual Recognition service. 

This valuable service helps with the process of accurately analyzing, classifying, and training images using machine learning logic (although, to be sure, it still requires reasonable amounts of relevant training data to begin with, but more on that in a bit). 

Thankfully, there is a set of built-in models that is available to us to provide highly accurate results without endless training. The models are as follows:

• General model: General classifier categories
• Face model: Locate faces within an image, gender, and age
• Explicit model: Whether an image is inappropriate for general use
• b model: Specifically for images of food items

In this chapter's project, we will show how to use the Visual Recognition Service and the Face model to build an end-to-end working solution that can look at an image of a human face, perform expression analysis and simple classification, and ultimately, determine whether the individual is feeling happy or sad.