Tag ranking based on salient region graph propagation
Lists
Only you see this
Abstract
In this paper, the problem of tag ranking by propagating relevance over community-contributed images and their associated tags is explored. To rank the tags more accurately, we propose a novel tag ranking approach based on the salient region. Firstly, we extract the salient region sub-image based on Itti model, and then construct two graphs with the whole image and the salient region sub-image. Secondly, we use the graph-based model to propagate the relevance of labels. Finally, we calculate the relevance score according to the results of the relevance propagation on the two sparse graphs. And then the new order of tags is determined by the relevance of the tags from high to low. Compared to existing methods, the proposed method considers not only the relationship between the whole images, but also the relationship between the salient regions. Therefore, it enhances the accuracy of the tag ranking. Experimental results conducted on a real dataset demonstrate that the ranking result of the proposed approach is closer to the manual rank.
BibTeX
@inproceedings{tang2015tag,
author = {Jinhui Tang and Minxian Li and Zechao Li and Chunxia Zhao},
title = {{Tag ranking based on salient region graph propagation}},
booktitle = {Multim. Syst},
year = {2015},
url = {https://openreview.net/forum?id=EK5Ybs8qmCL},
}Loading related content...