PCA is a standard dimensionality reduction technique. You will use SVD/PCA for facial recognition, following the Eigenfaces algorithm. Eigenfaces is a technique which treats images as vectors and finds a set of orthonormal basis images using SVD. The top basis images can be used to reconstruct images.
You are (as a group) to implement the Eigenfaces algorith. You can use any language to choose but must do the implementation your self. (no eigenfaces library) You will need to find/create a database of face images to be used as a training set. You will need a testing set of additional face images which contains both people in the original training set and not in the training set. You should report your findings and explore the relationship between the number of basis vectors and the recognition ability.
Your code will be submitted on canvas. Your code is due Monday 11/19 at 11:59pm. You should be prepared to talk about your project in class on 11/19.