In this homework, you will work on **pattern recognition problems** that require position invariance. Specifically, you will work on the problem of recognizing faces in images. In typical pictures of faces, the face is rarely perfectly centered. Different pictures of the same person may have the face shifted by varying amounts. The classifier must recognize the face regardless of this indeterminacy of position.  This calls for position-invariant models, specifically Convolutional Neural Networks, or CNNs.

Face classification is a **closed-set multiclass classification problem**, where the subjects in the test set have also been seen in the training set, although the precise pictures in the test set will not be in the training set. For this to achieve high accuracy it is only required that the embeddings for (all pictures of) the subjects in our "vocabulary" be linearly separable from each other.