This is a project of Computer Vision module at University of Southampton. Give a set of training images, we implement 3 ways to do classification:
1: Simple K-Nearest Neighbor on Tiny image feature 2: Linear Classification on Vector quantisation of K-Mean centres of block feature 3: Transfer learning
I uploaded the first two parts into thid repository, because the third one just be similar to the sample code from Tensorflow. I also uploaded the final report.
I did all the coding job by myself and completed report with teammates.