Cosine Multilinear Principal Component Analysis for Recognition
文献类型:期刊论文
作者 | Han, Feng1; Leng, Chengcai1; Li, Bing2; Basu, Anup3; Jiao, Licheng4 |
刊名 | IEEE TRANSACTIONS ON BIG DATA |
出版日期 | 2023-12-01 |
卷号 | 9期号:6页码:1620-1630 |
ISSN号 | 2332-7790 |
关键词 | Tensors Principal component analysis Mathematical models Linear programming Iterative methods Robustness Matrix decomposition Multilinear principal component analysis angle tensor analysis pattern recognition |
DOI | 10.1109/TBDATA.2023.3301389 |
通讯作者 | Leng, Chengcai(ccleng@nwu.edu.cn) |
英文摘要 | Existing two-dimensional principal component analysis methods can only handle second-order tensors (i.e., matrices). However, with the advancement of technology, tensors of order three and higher are gradually increasing. This brings new challenges to dimensionality reduction. Thus, a multilinear method called MPCA was proposed. Although MPCA can be applied to all tensors, using the square of the F-norm makes it very sensitive to outliers. Several two-dimensional methods, such as Angle 2DPCA, have good robustness but cannot be applied to all tensors. We extend the robust Angle 2DPCA method to a multilinear method and propose Cosine Multilinear Principal Component Analysis (CosMPCA) for tensor representation. Our CosMPCA method considers the relationship between the reconstruction error and projection scatter and selects the cosine metric. In addition, our method naturally uses the F-norm to reduce the impact of outliers. We introduce an iterative algorithm to solve CosMPCA. We provide detailed theoretical analysis in both the proposed method and the analysis of the algorithm. Experiments show that our method is robust to outliers and is suitable for tensors of any order. |
WOS关键词 | ROBUST TENSOR ANALYSIS ; FACE RECOGNITION ; 2DPCA ; REPRESENTATION ; NORM ; PCA ; MAXIMIZATION ; L1-NORM |
资助项目 | National Natural Science Foundation of China[61702251] ; Natural Science Basic Research Plan in Shaanxi Province of China[2018JM6030] ; Natural Sciences and Engineering Research Council of Canada ; Shaanxi Fundamental Science Research Project for Mathematics and Physics[22JSY010] |
WOS研究方向 | Computer Science |
语种 | 英语 |
出版者 | IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC |
WOS记录号 | WOS:001107490500017 |
资助机构 | National Natural Science Foundation of China ; Natural Science Basic Research Plan in Shaanxi Province of China ; Natural Sciences and Engineering Research Council of Canada ; Shaanxi Fundamental Science Research Project for Mathematics and Physics |
源URL | [http://ir.ia.ac.cn/handle/173211/55070] |
专题 | 多模态人工智能系统全国重点实验室 |
通讯作者 | Leng, Chengcai |
作者单位 | 1.Northwest Univ, Sch Math, Xian 710127, Shaanxi, Peoples R China 2.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 3.Univ Alberta, Dept Comp Sci, Edmonton, AB T6G 2E8, Canada 4.Xidian Univ, Sch Artificial Intelligence, Key Lab Intelligent Percept & Image Understanding, Minist Educ, Xian 710071, Shaanxi, Peoples R China |
推荐引用方式 GB/T 7714 | Han, Feng,Leng, Chengcai,Li, Bing,et al. Cosine Multilinear Principal Component Analysis for Recognition[J]. IEEE TRANSACTIONS ON BIG DATA,2023,9(6):1620-1630. |
APA | Han, Feng,Leng, Chengcai,Li, Bing,Basu, Anup,&Jiao, Licheng.(2023).Cosine Multilinear Principal Component Analysis for Recognition.IEEE TRANSACTIONS ON BIG DATA,9(6),1620-1630. |
MLA | Han, Feng,et al."Cosine Multilinear Principal Component Analysis for Recognition".IEEE TRANSACTIONS ON BIG DATA 9.6(2023):1620-1630. |
入库方式: OAI收割
来源:自动化研究所
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。