Efficient Re-Ranking of Images from the Web using Bag based Method

1. ABSTRACT An image retrieval system is a computer system for browsing, searching and retrieving images from a large database of digital images. Given a textual query in traditional text based image retrieval (TBIR),relevant images are to be re ranked using visual features after the initial text based image search. In this paper, we propose a new bag based re ranking framework for large scale TBIR. We compute this problem as Multiple Instance Learning and Generalized Multiple Instance (GMI) learning method. To address the ambiguities on the instance labels in the positive and negative bags we propose a GMI settings. Also the user log performs the operation of individual user interaction with the system which improves the performance of image retrieval.