Social media sharing websites like Flickr allow users to annotate pictures with free Labels, which significantly contribute to the development of the web picture retrieval and organization. Label-based picture search is an important method to find pictures contributed by social users in such social websites. However, how to make the top ranked result relevant and with diversity is challenging. In this paper, we propose a social re-ranking system for Label-based picture retrieval with the consideration of picture’s relevance and diversity. We aim at re-ranking pictures according to their visual information, semantic information and social clues. The initial results include pictures contributed by different social users. Usually each user contributes several pictures. First we sort these pictures by inter-user re-ranking. Users that have higher contribution to the given query rank higher. Then we sequentially implement intra-user re-ranking on the ranked user’s picture set, and only the most relevant picture from each user’s picture set is selected. These selected pictures compose the final retrieved results. We build an inverted index structure for the social picture dataset to accelerate the searching process. Experimental results on Flickr dataset show that our social re-ranking method is effective and efficient
Article Details
Unique Paper ID: 146035
Publication Volume & Issue: Volume 4, Issue 11
Page(s): 1674 - 1678
Article Preview & Download
Share This Article
Join our RMS
Conference Alert
NCSEM 2024
National Conference on Sustainable Engineering and Management - 2024