This blog has been set up to record the progress of my Final Year Project.
The project aim is to develop a prototype for automatic image annotation.
This prototype will use recently released dataset of 100 million images ‘YFCC100M’ by Yahoo! and Flickr. The dataset contains computed image features such as GIST, LIRE, AlexNet and SIFT as well as tags, metadata and geographical locations. The tags suggested for a user query image will be based on tags from images from the dataset, to which it is similar.