Filters








244 Hits in 5.4 sec

Matching Larger Image Areas for Unconstrained Face Identification

Jack Gaston, Ji Ming, Danny Crookes
2018 IEEE Transactions on Cybernetics  
When the two matching patches y k and x l are small, the NCC R(y k , x l ) defined above may or may not offer much advantage over other types of distance or similarity metrics that have been used for image  ...  In this paper, we demonstrate our new approach by using a simple metric, NCC, for comparing images.  ... 
doi:10.1109/tcyb.2018.2846579 pmid:29994697 fatcat:n5gkwf7pevf5zijpdlkofahsru

Graph-Laplacian PCA: Closed-Form Solution and Robustness

Bo Jiang, Chris Ding, Bio Luo, Jin Tang
2013 2013 IEEE Conference on Computer Vision and Pattern Recognition  
Principal Component Analysis (PCA) is a widely used to learn a low-dimensional representation. In many applications, both vector data X and graph data W are available.  ...  Laplacian embedding is widely used for embedding graph data. We propose a graph-Laplacian PCA (gLPCA) to learn a low dimensional representation of X that incorporates graph structures encoded in W .  ...  This work is supported by National Natural Science Foundation of China (No.61073116, 61211130309, 61202228) and U.S. National Science Foundation NSF-CCF-0917274, NSF-DMS-0915228.  ... 
doi:10.1109/cvpr.2013.448 dblp:conf/cvpr/JiangDLT13 fatcat:etrh4fkxwnacxlev6opa3iptkm

Soft vector quantization and the EM algorithm

Ethem Alpaydın
1998 Neural Networks  
norm and the cluster priors which we could not with the other approaches.  ...  We extend this relation to their training, showing that learning rules by these models to estimate the cluster centers can be seen as approximations to the expectation-maximization (EM) method as applied  ...  This is done by defining a cluster center for each cluster and measuring the distance between an input and the cluster center using an appropriate metric.  ... 
doi:10.1016/s0893-6080(97)00147-0 pmid:12662823 fatcat:ki4axnb6lvatrcnvz3uluhe4zm

Joint Featurewise Weighting and Lobal Structure Learning for Multi-view Subspace Clustering [article]

Shi-Xun Lina, Guo Zhongb, Ting Shu
2020 arXiv   pre-print
Multi-view clustering integrates multiple feature sets, which reveal distinct aspects of the data and provide complementary information to each other, to improve the clustering performance.  ...  To address the above issues, we propose a novel multi-view subspace clustering method via simultaneously assigning weights for different features and capturing local information of data in view-specific  ...  (7) use Euclidean distance computed on the original data to guide the local structure learning.  ... 
arXiv:2007.12829v1 fatcat:mbagkzkh6zekdmfvda4eiii5k4

Distance metric learning by minimal distance maximization

Yaoliang Yu, Jiayan Jiang, Liming Zhang
2011 Pattern Recognition  
Following a systematic analysis of the multi-class LDR problem in a unified framework, we propose a new algorithm, called minimal distance maximization (MDM), to address the non-robustness issue.  ...  The principle behind MDM is to maximize the minimal between-class distance in the output space.  ...  Acknowledgments The authors are supported by National Natural Science Foundation of China (60571052) and Shanghai Leading Academic Discipline Project (B112).  ... 
doi:10.1016/j.patcog.2010.09.019 fatcat:h3pbgaifx5chnceztozp54jcj4

NC-link: A New Linkage Method for Efficient Hierarchical Clustering of Large-Scale Data

Yongkweon Jeon, Jaeyoon Yoo, Jongsun Lee, Sungroh Yoon
2017 IEEE Access  
INDEX TERMS Clustering algorithm, data mining, machine learning. 5594 2169-3536 2017 IEEE. Translations and content mining are permitted for academic research only.  ...  A run of HC requires multiple iterations, each of which needs to compute and update the pairwise distances between all intermediate clusters.  ...  As a distance measure, we use the Euclidean distance or the L 2 norm denoted by || · || in this paper; depending on the application, a different type of distance may be employed.  ... 
doi:10.1109/access.2017.2690987 fatcat:yurib2tb7jhy7mo44mi72try2a

Image segmentation with patch-pair density priors

Xiaobai Liu, Jiashi Feng, Shuicheng Yan, Hai Jin
2010 Proceedings of the international conference on Multimedia - MM '10  
A simple yet efficient multi-task joint sparse representation model is presented to augment the patch-pair similarities by harnessing the newly discovered patch-pair density priors.  ...  Patch-pair density prior, multi-task joint sparse representation, unsupervised image segmentation, multi-label image annotation.  ...  two clusterings of data C and C ′ .  ... 
doi:10.1145/1873951.1873968 dblp:conf/mm/LiuFYJ10 fatcat:6epernmzlbbspeciwn6t62li3y

Robust Hashing for Multi-View Data: Jointly Learning Low-Rank Kernelized Similarity Consensus and Hash Functions [article]

Lin Wu, Yang Wang
2016 arXiv   pre-print
To learn robust hash functions, a latent low-rank kernel function is used to construct hash functions in order to accommodate linearly inseparable data.  ...  Learning hash functions/codes for similarity search over multi-view data is attracting increasing attention, where similar hash codes are assigned to the data objects characterizing consistently neighborhood  ...  . • We motivate the problem of robust hashing over multiview data with nonlinear data distribution, and propose to learn the robust hash functions and a low-rank kernelized similarity matrix shared by  ... 
arXiv:1611.05521v1 fatcat:vtnotaqms5dd5b4vklca4sevzu

A streamlined scRNA-Seq data analysis framework based on improved sparse subspace clustering

Jujuan Zhuang, Lingyu Cui, Tianqi Qu, Changjing Ren, Junlin Xu, Tianbao Li, Geng Tian, Jialiang Yang
2021 IEEE Access  
The proposed optimization model is solved by the Linearized Alternating Direction Method of Multipliers, in which data completion and spectral clustering are combined as a whole by mutual constraint.  ...  One advantage of single-cell RNA sequencing is its ability in revealing cell heterogeneity by cell clustering.  ...  ) min Q trace(Q T LQ)s.t.Q T Q = I , Q ∈ R N ×k (14) which can be solved by spectral clustering.  ... 
doi:10.1109/access.2021.3049807 fatcat:jkvzt5xwe5e5tfc6vxalx7iad4

Clustered Multi-Task Learning for Automatic Radar Target Recognition

Cong Li, Weimin Bao, Luping Xu, Hua Zhang
2017 Sensors  
In this paper, we propose a clustered multi-task learning, which can reveal and share the multi-task relationships for radar target recognition.  ...  classification [13], we propose a new classification method based on clustered multi-task learning theory.  ...  Figure 6 shows that the multi-task learning with a trace-norm regularization has the lowest recognition rate.  ... 
doi:10.3390/s17102218 pmid:28953267 pmcid:PMC5676668 fatcat:qq26hymeoncfzcvdhypr5f76ru

Fuzzy Granule Manifold Alignment Preserving Local Topology

Wei Li, Jianwu Xue, Yumin Chen, Xuebai Zhang, Chao Tang, Qiang Zhang, Yifang Gao
2020 IEEE Access  
A projection is learned that can map instances described by two types of features to a low-dimensional space.  ...  Next, the local topology around the fuzzy granular vector is introduced and the optimal local topology matching can be achieved by minimizing their Frobenius norm.  ...  For more information, see https://creativecommons.org/licenses/by/4.0/. This article has been accepted for publication in a future issue of this journal, but has not been fully edited.  ... 
doi:10.1109/access.2020.3027311 fatcat:7yi6xqkbdregvg5tntcmhmxw4a

Mining Behavioral Groups in Large Wireless LANs [article]

Wei-jen Hsu, Debojyoti Dutta, Ahmed Helmy
2007 arXiv   pre-print
We represent the data using location preference vectors, and utilize unsupervised learning (clustering) to classify trends in user behavior using novel similarity metrics.  ...  We discover multi-modal user behavior for more than 60% of the users, and there are hundreds of distinct groups with unique behavioral patterns in both campuses.  ...  We show that a singular-value decomposition (SVD) based scheme not only provides the best summary of the data, but also leads to a distance metric that is robust to noise and computationally efficient.  ... 
arXiv:cs/0606002v2 fatcat:tvggwiprvbcnnbzltrzkv7l7hq

Effective Multi-Query Expansions: Collaborative Deep Networks for Robust Landmark Retrieval

Yang Wang, Xuemin Lin, Lin Wu, Wenjie Zhang
2017 IEEE Transactions on Image Processing  
In this paper we propose a novel framework, namely multi-query expansions, to retrieve semantically robust landmarks by two steps.  ...  The learned deep network is further applied to generate the features for all the other photos, meanwhile resulting into a compact multi-query set within such space.  ...  Motivated by this, in this paper we propose a novel method, depicted in Fig. 2 , for a robust landmark retrieval through a novel paradigm based on multi-query expansions over the social media data set  ... 
doi:10.1109/tip.2017.2655449 pmid:28103558 fatcat:k6wsv2mjszes7kgegyrg4urtsa

Face Super Resolution: A Survey

Sithara Kanakaraj, V. K. Govindan, Saidalavi Kalady
2017 International Journal of Image Graphics and Signal Processing  
Learning based methods are by far the immensely used technique.  ...  Sparse representation techniques, Neighborhood-Embedding techniques, and Bayesian learning techniques are all different approaches to learning based methods.  ...  The shrinkage parameter q of the lq norm regularization term is automatically approximated from the data to learn the linear coefficients for FH.  ... 
doi:10.5815/ijigsp.2017.05.06 fatcat:yubsqll3rnbvpcs3hhxfjhncnq

Decomposition into low-rank plus additive matrices for background/foreground separation: A review for a comparative evaluation with a large-scale dataset

Thierry Bouwmans, Andrews Sobral, Sajid Javed, Soon Ki Jung, El-Hadi Zahzah
2017 Computer Science Review  
The most representative problem formulation is the Robust Principal Component Analysis (RPCA) solved via Principal Component Pursuit (PCP) which decomposes a data matrix in a low-rank matrix and a sparse  ...  Then, we examine carefully each method in each robust subspace learning/tracking frameworks with their decomposition, their loss functions, their optimization problem and their solvers.  ...  Thus, the miminization problem is the following one: min L,S ||L|| p Sp + λ||S|| lq subj A − L − S = 0 (36) By replacing the Schatten-p norm and a lq-norm by their expression, the miminization problem  ... 
doi:10.1016/j.cosrev.2016.11.001 fatcat:vdh7ic4n6zfkjlccnyiq74z5wu
« Previous Showing results 1 — 15 out of 244 results