作者: Xin Liu , Jing Zhang , Li Zhuo , Zhenwei Li
DOI: 10.1109/ICMEW.2014.6890529
关键词:
摘要: With the rapid development of social media, research about its application and organization has attracted considerable attention. Many media sharing websites allow users to tag uploaded images with tags. However, tags are usually uncontrolled, ambiguous, overly personalized. One most challenging issues is how match textual visual image rank accurately. Because a great deal stored transmitted in compressed format, an optimized ranking based on vocabulary domain proposed order improve accuracy further reduce time. First, low-resolution constructed from data. Then created after SIFT descriptor extracted. Finally, neighbor voting model utilized Experimental results show that optimization method can significantly time under ensuring