This was the second assignment of the course where we focus on image keypoints and their descriptors. We mainly focus on Bag of Visual Words SIFT, SURF and RANSAC.
We did a horse vs bike classification where the accuracy shot upto 94.5% with just traditional ML algorithms like Logistic Regression and LinearSVM.
We also did a multiclass classification using the same methods on CIFAR-10 dataset.
The Panorama directory will help you to make a panorama using multiple pictures of a place.
Please refer to the pdf for a detailed report.