Case study 1 - facial recognition

Our first case study will be to predict the labels for image data with a popular dataset called the Labeled Faces in the Wild dataset from the scikit-learn library. The dataset is called the Olivetti Face dataset and it comprises pictures of famous people's faces, with appropriate labels. Our task is that of facial recognition, a supervised machine learning model that is able to predict the name of the person given an image of their face.

Applications of facial recognition

Image processing and facial recognition are far-reaching. The ability to quickly discern people's faces from a crowd of people in video/images is vital for physical security as well as for giant social media companies. Search engines such as Google, with their image search capabilities, are using image recognition algorithms to match images and quantify similarities to a point where we can upload a photo of someone to get all other images of that same person.

The data

Let's start with loading...