Classifying face images with SVM
Finally, it is time to build an SVM-based face image classifier using everything you just learned. We will do so in parts, exploring the image dataset first.
Exploring the face image dataset
We will use the Labeled Faces in the Wild (LFW) people dataset (https://scikit-learn.org/stable/modules/generated/sklearn.datasets.fetch_lfw_people.html) from scikit-learn. It consists of more than 13,000 curated face images of more than 5,000 famous people. Each class has various numbers of image samples.
First, we load the face image data as follows:
>>> from sklearn.datasets import fetch_lfw_people
Downloading LFW metadata: https://ndownloader.figshare.com/files/5976012
Downloading LFW metadata: https://ndownloader.figshare.com/files/5976009
Downloading LFW metadata: https://ndownloader.figshare.com/files/5976006
Downloading LFW data (~200MB): https://ndownloader.figshare.com/files/5976015
>>> face_data = fetch_lfw_people(min_faces_per_person...