Vectorize your code. Non$-$optimized code may be penalized.

Deep Features for Image Classification:

$($a$)$ In this assignment, you will learn to use a pre$-$trained deep neural network. You are given a set of training images and test images. Load a pre$-$trained model of your choice $($ResNet$,$ AlexNet, etc.$)$ and extract deep features for every image before the last classification head of the model $($For example, if the model is ResNet, the classification head $-$ generally named as model.fc$,$ maps the last layer features in $R^{2048}$ to $R^{1000}-1000$ classes. Here, you should extract the features in $R^{2048}).$ Then, compute the image classification accuracy for the test set using the KNN classifier