Matrix completion by Truncated Nuclear Norm Regularization | |
Zhang, Debing ; Hu, Yao ; Ye, Jieping ; Li, Xuelong ; He, Xiaofei | |
2012 | |
会议名称 | 2012 ieee conference on computer vision and pattern recognition, cvpr 2012 |
会议日期 | june 16, 2012 - june 21, 2012 |
会议地点 | providence, ri, united states |
页码 | 2192-2199 |
英文摘要 | estimating missing values in visual data is a challenging problem in computer vision, which can be considered as a low rank matrix approximation problem. most of the recent studies use the nuclear norm as a convex relaxation of the rank operator. however, by minimizing the nuclear norm, all the singular values are simultaneously minimized, and thus the rank can not be well approximated in practice. in this paper, we propose a novel matrix completion algorithm based on the truncated nuclear norm regularization (tnnr)by only minimizing the smallest n-r singular values, where n is the number of singular values and r is the rank of the matrix. in this way, the rank of the matrix can be better approximated than the nuclear norm. we further develop an efficient iterative procedure to solve the optimization problem by using the alternating direction method of multipliers and the accelerated proximal gradient line search method. experimental results in a wide range of applications demonstrate the effectiveness of our proposed approach. |
收录类别 | CPCI(ISTP) ; EI |
产权排序 | 3 |
会议录 | 2012 ieee conference on computer vision and pattern recognition, cvpr 2012 |
会议录出版者 | ieee computer society, 2001 l street n.w., suite 700, washington, dc 20036-4928, united states |
会议录出版地 | united states |
语种 | 英语 |
ISSN号 | 10636919 |
ISBN号 | 9781467312264 |
内容类型 | 会议论文 |
源URL | [http://ir.opt.ac.cn/handle/181661/20536] |
专题 | 西安光学精密机械研究所_光学影像学习与分析中心 |
推荐引用方式 GB/T 7714 | Zhang, Debing,Hu, Yao,Ye, Jieping,et al. Matrix completion by Truncated Nuclear Norm Regularization[C]. 见:2012 ieee conference on computer vision and pattern recognition, cvpr 2012. providence, ri, united states. june 16, 2012 - june 21, 2012. |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
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。
修改评论