Distance metric learning for pattern recognition.

A good distance metric is crucial for unsupervised learning from high-dimensional data. ... In this paper, we propose a novel unsupervised Adaptive Metric Learning algorithm, called AML, which performs clustering and distance metric learning simultaneously. ... Acknowledgments This research is sponsored by the Center for Evolutionary Functional Genomics of the Biodesign Institute at ASU and by the National Science Foundation Grant IIS-0612069. ...

doi:10.1109/cvpr.2007.383103 dblp:conf/cvpr/YeZL07 fatcat:h2nfpum7bbfvri5zvxpthwfgli

Given must-link and cannot-link information, our goal is to learn a Mahalanobis distance metric. ... Distance metric is a key issue in many machine learning algorithms. This paper considers a general problem of learning from pairwise constraints in the form of must-links and cannot-links. ... Application to face pose estimation Face recognition is a challenging research direction in pattern recognition. ...

doi:10.1016/j.patcog.2008.05.018 fatcat:zglrulttxjhexluc6owghyqq44

To effectively explore the log data, we propose a novel semi-supervised distance metric learning technique, called "Laplacian Regularized Metric Learning" (LRML), for learning robust distance metrics for ... We show that reliable metrics can be learned from real log data even they may be noisy and limited at the beginning stage of a CIR system. ... In this paper, we propose a new semi-supervised distance metric learning scheme for incorporating unlabeled data in the distance metric learning task. ...

doi:10.1109/cvpr.2008.4587351 dblp:conf/cvpr/HoiLC08 fatcat:d3jmrgmhejhcpj3nb5y6uka4pm

Compared with Euclidean distance metric, our S-NR distance metric can further jointly reduce the intra-class distances and enlarge the inter-class distances for learned features. ... between image pairs, which well justify its suitability for deep metric learning. ... Related Work Metric Learning Metric learning methods, which have been widely applied to image retrieval, clustering and recognition tasks, have attracted much attention. ...

doi:10.1109/cvpr.2019.00495 dblp:conf/cvpr/YuanDTTC19 fatcat:ueug7mysj5ccxcygqiy3xhtbo4

For the NCM classifier we introduce a new metric learning approach, and we also introduce an extension to allow for richer class representations. ... Since the performance of distance-based classifiers heavily depends on the used distance function, we cast the problem into one of learning a low-rank metric, which is shared across all classes. ... Fig. 7 : 7 The nearest classes for two reference classes using the the ℓ 2 distance and metric learned by NCMML. ...

doi:10.1007/978-1-4471-5520-1_9 dblp:series/acvpr/MensinkVP13 fatcat:ntz7nljgrbdlxbyuss4f3dbdee

Relevant Component Analysis (RCA) has been proposed for learning distance metrics with contextual constraints for image retrieval. However, RCA has two important disadvantages. ... Compared with other complicated methods for distance metric learning, our algorithms are rather simple to understand and very easy to solve. ... Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06) Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition ...

doi:10.1109/cvpr.2006.167 dblp:conf/cvpr/HoiLLM06 fatcat:wdmcdhmuxbd3njjqp73in77lye

The FHLSNN pattern classifier is evaluated for generalization performance under recognition rate, training time and testing time. ... This analysis will help to select a suitable distance metric for fuzzy neural network classifier for particular application. ... As patterns given for learning of classifier are totally different for the patterns given for testing hence 100% recognition cannot be expected since testing is for generalization ability of classifiers ...

doi:10.5120/16612-6450 fatcat:gxulmztfrregnb7bc3h5u7vdmq

Behavior recognition is an attractive direction in the computer vision domain. In this paper, we propose a novel behavior recognition method based on prototype learning using metric learning. ... And the metric learning algorithm is used to advance the performance of the prototype learning. ... Call the Eric algorithms for the distance learning. 4. Choose a pattern x n from the pattern pool. 5. ...

doi:10.1109/icpr.2010.638 dblp:conf/icpr/ZhuHYL10 fatcat:boqrwjgehfdxdphl57gixhjvgi

However, the role of the descriptor can differ with different classifiers and distance metrics for diverse issues in face recognition. ... Euclidian, Manhattan, Minkowski, G-statistics and chisquare dissimilarity metrics to illustrate differences in performance with respect to assorted issues in face recognition using six benchmark databases ... metrics for face recognition with pose variant images. ...

doi:10.1590/1678-4324-2016161057 fatcat:mkwuawrb2ne63bkp6plkcbxhe4

DOAJ SciELO

It uses an adaptation of the Hausdorff distance to determine the similarity between an input pattern and a learned representation. ... The system uses the HAusdorf-Voronoi NETwork (HAVNET), an artificial neural network designed for two-dimensional binary pattern recognition. ... The Hausdorff distance exhibits many desirable properties for pattern recognition. (a) It is known to be a metric over the set of all closed bounded sets. ...

doi:10.1117/12.382901 fatcat:joi2qt2psrb57c2inaxcp4nvw4

However, visual recognition is broader than just pair-matching: what we learn and how we learn it has important implications for e↵ective algorithms. ... By studying these violations, useful insights come to light: we make the case for local distances and systems that leverage outside information to solve the general recognition problem. ... This is consistent with supporting prior work [14] in pattern recognition that shows increasing discriminative power for non-metric distance measures over visual data. ...

doi:10.1016/j.patcog.2014.02.018 fatcat:ynrs6motbjdipmrm5bzu2u6xuu

Chang and colleagues propose a non-parametric method for learning global metrics for content-based image retrieval. ... The challenge of automatic pattern recognition is to develop computational methods which learn to distinguish among a number of classes represented by examples. ...

doi:10.1016/j.patcog.2006.04.004 fatcat:hb57tnpkgrfudjd4aryqbosrby

In this paper, we propose to learn the best distance for the k-nearest neighbor (k-NN) classifier. ... The work presented here can open new research paths towards a novel automatic musical symbols recognition module for handwritten scores. 10th International Conference on Machine Learning and Applications ... Supervised distance metric learning can be further divided into global distance metric learning and local distance metric learning. ...

doi:10.1109/icmla.2011.94 dblp:conf/icmla/RebeloTSC11 fatcat:6ejxx75fl5dfjcda7sokeehdpa

DISTANCE METRIC LEARNING Distance metric learning is an emerging subarea of machine learning in which the underlying metric is itself adapted to improve the results of classification and pattern recognition ... METRIC LEARNING FOR NEAREST-NEIGHBOR CLASSIFICATION Convex Optimizations for Distance Metric Learning and Pattern Classification by the majority label of their kNN in the training set. ... Looking forward, we anticipate many such applications given the ubiquitous role of distance metrics in both nonparametric and parametric models of classification. ...

doi:10.1109/msp.2010.936013 fatcat:ffx2slovabbelmoil7nz2iptia

To address this impediment of manifold learning, we investigated the combination of manifold learning and distance metric learning for the generation of a representation that is both discriminative and ... We demonstrate that this approach is effective for multi view face recognition. ... The proposed scheme can be useful for face recognition and other applications, and may be extended to different combinations of distance metric learning and manifold learning approaches. ...

doi:10.1109/icpr.2010.165 dblp:conf/icpr/TsagkatakisS10 fatcat:zlixruu6ijbl7bmd2lzqi2ijzu

Adaptive Distance Metric Learning for Clustering

Preserved Fulltext

Learning a Mahalanobis distance metric for data clustering and classification

Preserved Fulltext

Semi-supervised distance metric learning for Collaborative Image Retrieval

Preserved Fulltext

Signal-To-Noise Ratio: A Robust Distance Metric for Deep Metric Learning

Preserved Fulltext

Large Scale Metric Learning for Distance-Based Image Classification on Open Ended Data Sets [chapter]

Preserved Fulltext

Learning Distance Metrics with Contextual Constraints for Image Retrieval

Preserved Fulltext

Fuzzy Hyperline Segment Neural Network Pattern Classifier with Different Distance Metrics

Preserved Fulltext

Prototype Learning Using Metric Learning Based Behavior Recognition

Preserved Fulltext

An Empirical Evaluation of the Local Texture Description Framework-Based Modified Local Directional Number Pattern with Various Classifiers for Face Recognition

Preserved Fulltext

Sign language recognition using competitive learning in the HAVNET neural network

Preserved Fulltext

Good recognition is non-metric

Preserved Fulltext

Similarity-based pattern recognition

Preserved Fulltext

Metric Learning for Music Symbol Recognition

Preserved Fulltext

Convex Optimizations for Distance Metric Learning and Pattern Classification [Applications Corner

Preserved Fulltext

Manifold Modeling with Learned Distance in Random Projection Space for Face Recognition

Preserved Fulltext