文献类型：期刊文献

英文题名：Duplicate Image Representation Based on Semi-Supervised Learning

作者：Chen, Ming[1];Yan, Jinghua[2,3];Gao, Tieliang[4];Li, Yuhua[1];Ma, Huan[1]

第一作者：Chen, Ming

通讯作者：Chen, M[1]

机构：[1]Zhengzhou Univ Light Ind, Software Engn Coll, Zhengzhou, Peoples R China;[2]Natl Comp Network Emergency Response Tech Team, Beijing, Peoples R China;[3]Coordinat Ctr China, Beijing, Peoples R China;[4]Xinxiang Univ, Sch Business, Xinxiang, Henan, Peoples R China

第一机构：Zhengzhou Univ Light Ind, Software Engn Coll, Zhengzhou, Peoples R China

通讯机构：[1]corresponding author), Zhengzhou Univ Light Ind, Software Engn Coll, Zhengzhou, Peoples R China.

年份：2022

卷号：14

期号：1

外文期刊名：INTERNATIONAL JOURNAL OF GRID AND HIGH PERFORMANCE COMPUTING

收录：EI(收录号：20231113698416);Scopus(收录号：2-s2.0-85149682440);WOS:【ESCI(收录号:WOS:000916579600020)】；

语种：英文

外文关键词：BoF Model; Duplicate Image Detection; Metric Similarity; Real-Time Retrieval; Semantic Similarity; Semi-Supervised Learning; Unsupervised Learning; Visual Dictionary

摘要：For duplicate image detection, the more advanced large-scale image retrieval systems in recent years have mainly used the bag-of-feature (BoF) model to meet the real-time. However, due to the lack of semantic information in the training process of the visual dictionary, BoF model cannot guarantee semantic similarity. Therefore, this paper proposes a duplicate image representation algorithm based on semi-supervised learning. This algorithm first generates semi-supervised hashes and then maps the image local descriptors to binary codes based on semi-supervised learning. Finally, an image is represented by a frequency histogram of binary codes. Since the semantic information can be effectively introduced through the construction of the marker matrix and the classification matrix during the training process, semi-supervised learning can guarantee the metric similarity of the local descriptors and also guarantee the semantic similarity. And the experimental results also show this algorithm has a better retrieval effect compared with traditional algorithms.