中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 计算技术研究所
  3. 中国科学院计算技术研究所
  4. 中国科学院计算技术研究所期刊论文
  5. 英文
A two-stage hybrid probabilistic topic model for refining image annotation

文献类型：期刊论文

作者Tian, Dongping1; Shi, Zhongzhi2
刊名INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
出版日期2020-02-01
卷号11期号:2页码:417-431
关键词Refining image annotation Semantic gap Expectation-maximization PLSA Max-bisection Image retrieval
ISSN号1868-8071
DOI10.1007/s13042-019-00983-w
英文摘要Refining image annotation has become one of the core research topics in computer vision and pattern recognition due to its great potentials in image retrieval. However, it is still in its infancy and is not sophisticated enough to extract perfect semantic concepts just according to the image low-level features. In this paper, we propose a two-stage hybrid probabilistic topic model to improve the quality of automatic image annotation. To start with, a probabilistic latent semantic analysis model with asymmetric modalities is learned to estimate the posterior probabilities of each annotation keyword, during which the image-to-word relation can be well established. Next, a label similarity graph is constructed by a weighted linear combination of label similarity and visual similarity of images associated with the corresponding labels. By this way, the information from image low-level visual features and high-level semantic concepts can be seamlessly integrated by fully taking into account the word-to-word and image-to-image relations. Finally, the rank-two relaxation heuristics is exploited to further mine the correlation of the candidate annotations so as to capture the refining results, which plays a critical role in semantic based image retrieval. Extensive experiments show that the proposed model achieves not only superior annotation accuracy but also better retrieval performance.
资助项目National Program on Key Basic Research Project (973 Program)[2013CB329502] ; National Natural Science Foundation of China[61035003] ; National Natural Science Foundation of China[61202212] ; Tianchenghuizhi Fund for Innovation and Promotion of Education[2018A03036] ; Key R&D Program of the Shaanxi Province of China[2018GY-037]
WOS研究方向Computer Science
语种英语
WOS记录号WOS:000512019400011
出版者SPRINGER HEIDELBERG
源URL[http://119.78.100.204/handle/2XEOYT63/14775]  
专题中国科学院计算技术研究所期刊论文_英文
通讯作者Tian, Dongping
作者单位1.Baoji Univ Arts & Sci, Inst Comp Software, Baoji 721007, Shaanxi, Peoples R China
2.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100190, Peoples R China
推荐引用方式
GB/T 7714		 Tian, Dongping,Shi, Zhongzhi. A two-stage hybrid probabilistic topic model for refining image annotation[J]. INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS,2020,11(2):417-431.
APA Tian, Dongping,&Shi, Zhongzhi.(2020).A two-stage hybrid probabilistic topic model for refining image annotation.INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS,11(2),417-431.
MLA Tian, Dongping,et al."A two-stage hybrid probabilistic topic model for refining image annotation".INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS 11.2(2020):417-431.

入库方式： OAI收割

来源：计算技术研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。