Computationally modeling visual aesthetics

Abstract

With the growth of the internet and e-commerce, many people now shopped online. Images are being shared on the internet every single second and have caused an exponential growth in the number of images uploaded online. Users now look out for speed, web pages with less navigation or clicks and programs for instance, online blog shops that are efficient to cater to their different needs in this busy paced world that we live in today. In this project, the aim is to develop an image judging system that will pick up images that are visually appealing to save user’s time going through redundant images through understanding feature extraction methods such as Scale-Invariant Feature Transform (SIFT), dense SIFT and convolutional neural network for feature extraction with the combination of Rank SVM to model visual aesthetic.