Visual Reranking with Improved Image Graph

PDF

This paper introduces an improved reranking method for the Bag-of-Words (BoW) based image search. Built on [1], a directed image graph robust to outlier distraction is proposed. In our approach, the relevance among images is encoded in the image graph, based on which the initial rank list is refined. Moreover, we show that the rank-level feature fusion can be adopted in this reranking method as well. Taking advantage of the complementary nature of various features, the reranking performance is further enhanced. Particularly, we exploit the reranking method combining the BoW and color information. Experiments on two benchmark datasets demonstrate that ourmethod yields significant improvements and the reranking results are competitive to the state-of-the-art methods.
Submitted 3 Jun 2014 to Computer Vision and Pattern Recognition [cs.CV]
Published 4 Jun 2014
http://arxiv.org/abs/1406.0680
http://arxiv.org/pdf/1406.0680.pdf

0 comments