基于潜在主题融合的跨媒体图像语义标注

doi:10.3969/j.issn.0372-2112.2014.05.024

电子学报

基于潜在主题融合的跨媒体图像语义标注

刘杰^1,2, 杜军平¹

1. 北京邮电大学计算机学院, 北京 100876;
2. 中国电子科技集团公司第三十研究所, 四川成都 610041

收稿日期:2013-01-07 修回日期:2013-06-17 出版日期:2014-05-25
- 通讯作者:
- 杜军平
- 作者简介:
- 刘杰 男，1984年出生，博士，工程师，中国电子科技集团公司第三十研究所，主要研究方向：智能信息处理、机器学习、网络通信. E-mail：sleetext2@163.com
- 基金资助:
- 国家973重点基础研究发展计划 (No.2012CB821200，No.2012CB821206); 国家自然科学基金 (No.61320106006)

Latent Topic Fusion-Based Cross-Media Image Semantic Annotation

LIU Jie^1,2, DU Jun-ping¹

1. School of Computer, Beijing University of Posts and Telecommunications, Beijing 100876, China;
2. No.30 Institute of China Electronics Technology Group Corporation, Chengdu, Sichuan 610041, China

Received:2013-01-07 Revised:2013-06-17 Online:2014-05-25 Published:2014-05-25
- Supported by:
- National Program on Key Basic Research Project of China (973 Program) (No.2012CB821200, No.2012CB821206); National Natural Science Foundation of China (No.61320106006)

摘要/Abstract

摘要： 图像语义标注是图像语义分析研究中的一个重要问题.在主题模型的基础上，本文提出一种新颖的跨媒体图像标注方法来进行图像间语义的传播.首先，对训练图像使用主题模型，抽取视觉模态和文本模态信息的潜在语义主题.然后，通过使用一个权重参数来融合两种模态信息的主题分布，从而学习到一种融合主题分布.最后，在融合主题分布的基础上训练一个标注模型来给目标图像赋予合适的语义信息.在标准的MSRC和Corel5K数据集上将提出的方法与最近著名的标注方法进行比较实验.标注性能的详细评价结果表明提出方法的有效性.

关键词: 图像语义标注, 跨媒体, 主题模型, 加权融合

Abstract: Image semantic annotation is an important issue in image semantic analysis research.Based on the topic model,this paper proposes a novel cross-media image annotation approach for propagating the semantics among images.First,the topic model is used to capture the latent semantic topics from the visual and textual modal information in the training images.Then,a fused topic distribution is learned by merging the topic distribution of each modality using a weight parameter.Finally,an annotation model based on the fused topic distribution is trained to assign the target images using appropriate semantics.A comparison of the proposed approach with the recent state-of-the-art annotation approaches on the standard MSRC and Corel5K datasets is presented,and a detailed evaluation of the performance shows the validity of our approach.

Key words: image semantic annotation, cross media, topic model, topic-weighted fusion

中图分类号:

TP37
TP391.4

刘杰, 杜军平. 基于潜在主题融合的跨媒体图像语义标注[J]. 电子学报, DOI: 10.3969/j.issn.0372-2112.2014.05.024.

LIU Jie, DU Jun-ping . Latent Topic Fusion-Based Cross-Media Image Semantic Annotation[J]. Acta Electronica Sinica, DOI: 10.3969/j.issn.0372-2112.2014.05.024.

参考文献

[1] A Tousch,S Herbin,J Audibert.Semantic hierarchies for image annotation:A survey[J].Pattern Recognition,2012,45(1):333-345.
[2] J Tang,Z Zha,D Tao.Semantic-gap-oriented active learning for multi-label image annotation[J].IEEE Transactions on Image Processing,2012,21(4):2354-2360.
[3] D S Zhang,Md M Islam,G J Lu.A review on automatic image annotation techniques[J].Pattern Recognition,2012,45(1):346-362.
[4] A Makadia,V Pavlovic,S Kumar.Baselines for image annotation[J].International Journal of Computer Vision,2010,90(1):88-105.
[5] S Feng,R Manmatha,V Lavrenko.Multiple Bernoulli relevance models for image and video annotation[A].In Proc.IEEE Conf.on Computer Vision and Pattern Recognition[C].Washington,DC,USA:IEEE,2004:1002-1009.
[6] Z W Lu,H S I Horace.Automatic image annotation based on generalized relevance models[J].Journal of Signal Processing Systems,2011,65(1):23-33.
[7] J Zhong,Q G Sun,X Li,L S Wen.A novel feature selection method based on probability latent semantic analysis for Chinese text classification[J].Chinese Journal of Electronics,2011,20(2):228-232.
[8] X Ke,S Z Li,D L Cao.A two-level model for automatic image annotation[J].Multimedia Tools and Applications,2012,61(1):195-212.
[9] F Monay,D G Perez.Modeling semantic aspects for cross-media image indexing[J].IEEE Trans.Pattern Analysis and Machine Intelligence,2007,29(10):1802-1817.
[10] D M Blei.Probabilistic topic models[J].Communications of the ACM,2012,55(4):77-84.
[11] D Putthividhy,H T Attias,S S Nagarajan.Topic regression multi-modal latent Dirichlet allocation for image annotation[A].In Proc.IEEE Conf.on Computer Vision and Pattern Recognition[C].La Jolla,CA,USA:IEEE,2010:3408-3415.
[12] H Bay,A Eelaars,L V Gool.Speed-up robust features (SURF)[J].Computer Vision and Image Understanding,2008,110(3):346-359.
[13] Y Jiang,R Wang,P Zhang.Texture description based on multiresolution moments of image histograms[J].Optical Engineering,2008,47(3):037005.
[14] J Liu,J P Du,X R Wang.Research on the robust image representation scheme for natural scene categorization[J].Chinese Journal of Electronics,2013,22(2):341-346. [LL]
[15] W J Wen,D Xu,Y J Tang,S Y Liu,S H Feng.Mutual information based codebooks construction for natural scene categorization[J].Chinese Journal of Electronics,2011,20(3):419-424.
[16] M B Christopher.Pattern Recognition and Machine Learning[M].New York,USA.Springer.2006.
[17] J Shotton,J Winn,C Rother,A Criminisi.Textonboost:Joint appearance,shape and context modeling for multi-class object recognition and segmentation[A].In Proc 9th European Conf.on Computer Vision[C].Graz,Austria:Elsevier,2006:1-15.
[18] P Duygulu,K Barnard,J F G de Freitas,D A Forsyth.Object recognition as machine translation:Learning a lexicon for a fixed image vocabulary[J].Lecture Notes in Computer Science,2006,2353:349-354.

基于潜在主题融合的跨媒体图像语义标注

Latent Topic Fusion-Based Cross-Media Image Semantic Annotation

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 14

编辑推荐

Metrics

[1]	吴绿, 张馨月, 唐茉, 王梓, 王永安. Focus+Context语义表征的场景图像分割[J]. 电子学报, 2021, 49(3): 596-604.
[2]	唐焕玲, 郑涵, 刘艳红, 马思源, 窦全胜, 鲁明羽. Tr-SLDA:一种面向交叉领域的迁移主题模型[J]. 电子学报, 2021, 49(3): 605-613.
[3]	李志欣, 凌锋, 张灿龙, 马慧芳. 融合两级相似度的跨媒体图像文本检索[J]. 电子学报, 2021, 49(2): 268-274.
[4]	唐焕玲, 窦全胜, 于立萍, 宋英杰, 鲁明羽. 有监督主题模型的SLDA-TC文本分类新方法[J]. 电子学报, 2019, 47(6): 1300-1308.
[5]	张雄, 陈福才, 黄瑞阳. 基于双词主题模型的半监督实体消歧方法研究[J]. 电子学报, 2018, 46(3): 607-613.
[6]	孙小兵, 刘湘月, 李斌, 张伟佳. 基于相关主题模型的程序网络自动构建与分析[J]. 电子学报, 2017, 45(5): 1052-1056.
[7]	胡振涛, 胡玉梅, 刘先省. 量测提升卡尔曼滤波[J]. 电子学报, 2016, 44(5): 1149-1155.
[8]	吕品, 汪鑫, 罗宜元, 计春雷. 基于主题模型的(Aspect,Rating)摘要生成方法研究[J]. 电子学报, 2016, 44(12): 3036-3043.
[9]	田刚, 何克清, 王健, 孙承爱, 徐建建. 面向领域标签辅助的服务聚类方法[J]. 电子学报, 2015, 43(7): 1266-1274.
[10]	丁宇新, 燕泽权, 冯威, 薛成龙, 周迪. 基于有监督主题模型的排序学习算法[J]. 电子学报, 2015, 43(2): 333-337.
[11]	叶程, 刘真, 吴明光. 基于球坐标及三角形插值的颜色信号色域映射算法[J]. 电子学报, 2015, 43(11): 2180-2186.
[12]	张茜, 刘真, 张建青. 跨媒体复现中CIECAM02背景预测性能评价[J]. 电子学报, 2015, 43(10): 1918-1923.
[13]	李晓旭, 李睿凡, 冯方向, 曹洁, 王小捷. 多视图有监督的LDA模型[J]. 电子学报, 2014, 42(10): 2040-2044.
[14]	王李冬, 魏宝刚, 袁杰. 基于概率主题模型的文档聚类[J]. 电子学报, 2012, 40(11): 2346-2350.