SIA OpenIR  > 机器人学研究室
Self-Supervised Online Metric Learning With Low Rank Constraint for Scene Categorization
Cong Y(丛杨); Liu J(刘霁); Yuan JS(袁浚菘); Luo JB(罗杰波)
Indexed BySCI ; EI
EI Accession number20132516440364
WOS IDWOS:000321926600022
Contribution Rank1
Funding OrganizationNatural Science Foundation of China [61105013]; Nanyang Assistant Professorship [M4080134]; NTU CoE Seed [M4081039]
KeywordLow Rank Online Learning Metric Learning Semi-supervised Learning Scene Categorization
AbstractConventional visual recognition systems usually train an image classifier in a bath mode with all training data provided in advance. However, in many practical applications, only a small amount of training samples are available in the beginning and many more would come sequentially during online recognition. Because the image data characteristics could change over time, it is important for the classifier to adapt to the new data incrementally. In this paper, we present an online metric learning method to address the online scene recognition problem via adaptive similarity measurement. Given a number of labeled data followed by a sequential input of unseen testing samples, the similarity metric is learned to maximize the margin of the distance among different classes of samples. By considering the low rank constraint, our online metric learning model not only can provide competitive performance compared with the state-of-the-art methods, but also guarantees convergence. A bi-linear graph is also defined to model the pair-wise similarity, and an unseen sample is labeled depending on the graph-based label propagation, while the model can also self-update using the more confident new samples. With the ability of online learning, our methodology can well handle the large-scale streaming video data with the ability of incremental self-updating. We evaluate our model to online scene categorization and experiments on various benchmark datasets and comparisons with state-of-the-art methods demonstrate the effectiveness and efficiency of our algorithm.
WOS HeadingsScience & Technology ; Technology
WOS SubjectComputer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS Research AreaComputer Science ; Engineering
Citation statistics
Document Type期刊论文
Corresponding AuthorCong Y(丛杨)
Affiliation1.State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, China
2.School of Electrical and Electronics Engineering, Nanyang Technological University, 639798 Singapore, Singapore
3.Department of Computer Science, University of Wisconsin, Wisconsin, MA 53706, United States
4.Department of Computer Science, University of Rochester, Rochester, NY 14627, United States
Recommended Citation
GB/T 7714
Cong Y,Liu J,Yuan JS,et al. Self-Supervised Online Metric Learning With Low Rank Constraint for Scene Categorization[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2013,22(8):3179-3191.
APA Cong Y,Liu J,Yuan JS,&Luo JB.(2013).Self-Supervised Online Metric Learning With Low Rank Constraint for Scene Categorization.IEEE TRANSACTIONS ON IMAGE PROCESSING,22(8),3179-3191.
MLA Cong Y,et al."Self-Supervised Online Metric Learning With Low Rank Constraint for Scene Categorization".IEEE TRANSACTIONS ON IMAGE PROCESSING 22.8(2013):3179-3191.
Files in This Item: Download All
File Name/Size DocType Version Access License
Self-Supervised Onli(1859KB)期刊论文出版稿开放获取ODC PDDLView Download
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Cong Y(丛杨)]'s Articles
[Liu J(刘霁)]'s Articles
[Yuan JS(袁浚菘)]'s Articles
Baidu academic
Similar articles in Baidu academic
[Cong Y(丛杨)]'s Articles
[Liu J(刘霁)]'s Articles
[Yuan JS(袁浚菘)]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Cong Y(丛杨)]'s Articles
[Liu J(刘霁)]'s Articles
[Yuan JS(袁浚菘)]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: Self-Supervised Online Metric Learning With Low Rank Constraint for Scene Categorization.pdf
Format: Adobe PDF
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.