Filters








21,877 Hits in 9.1 sec

Pairwise-Covariance Linear Discriminant Analysis

Deguang Kong, Chris Ding
2014 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
The new perspective also provides a natural way to properly weigh different pairwise distances, which emphasizes the pairs of class with small distances, and this leads to the proposed pairwise covariance  ...  In machine learning, linear discriminant analysis (LDA) is a popular dimension reduction method. In this paper, we first provide a new perspective of LDA from an information theory perspective.  ...  This research is partially supported by NSF-CCF-0917274 and NSF-DMS-0915228 grants.  ... 
doi:10.1609/aaai.v28i1.9008 fatcat:3cxbbwl5v5eszdinh7umqicrsy

Mixed Region Covariance Discriminative Learning for Image Classification on Riemannian Manifolds

Xi Liu, Zhengming Ma, Guo Niu
2019 Mathematical Problems in Engineering  
We describe a new covariance descriptor, which could improve the discriminative learning ability of region covariance descriptor by taking into account the mean of feature vectors.  ...  In this paper, we propose a subspace projection framework for the classification task on Riemannian manifolds and give the mathematical derivation for it.  ...  Based on the related works, we propose a discriminative learning method, Mixed Region Covariance Discriminative Learning (MRCDL) for image classification.  ... 
doi:10.1155/2019/1261398 fatcat:s6sn5wypsnb2hoedthukh7v4na

Covariate Shift Adaptation for Discriminative 3D Pose Estimation

Makoto Yamada, Leonid Sigal, Michalis Raptis
2014 IEEE Transactions on Pattern Analysis and Machine Intelligence  
In presence of training set bias, the learning results in a biased model whose performance degrades on the (target) test set.  ...  Under the assumption of covariate shift we propose an unsupervised domain adaptation approach to address this problem.  ...  Mainly, it is computationally efficient and can naturally control the adaptiveness to the test distribution.  ... 
doi:10.1109/tpami.2013.123 pmid:24356346 fatcat:2muodcw7rvcmvbeebs4yiphree

Discriminative Sparse Inverse Covariance Matrix: Application in Brain Functional Network Classification

Luping Zhou, Lei Wang, Philip Ogunbona
2014 2014 IEEE Conference on Computer Vision and Pattern Recognition  
In this paper, we propose a learning framework to effectively improve the discriminative power of SICEs by taking advantage of the samples in the opposite class.  ...  The proposed framework is applied to analyzing the brain metabolic covariant networks built upon FDG-PET images for the prediction of the Alzheimer's disease, and shows significant improvement of classification  ...  In this way, we can efficiently solve our discriminative learning problems.  ... 
doi:10.1109/cvpr.2014.396 dblp:conf/cvpr/ZhouWO14 fatcat:75qp73ahfvdpfm2ligwuxtau7u

No Bias Left behind: Covariate Shift Adaptation for Discriminative 3D Pose Estimation [chapter]

Makoto Yamada, Leonid Sigal, Michalis Raptis
2012 Lecture Notes in Computer Science  
In presence of training set bias, the learning results in a biased model whose performance degrades on the (target) test set.  ...  Under the assumption of covariate shift we propose an unsupervised domain adaptation approach to address this problem.  ...  Conclusions: We propose a simple, yet effective, unsupervised method for addressing training set bias through covariate shift adaptation in (structured) prediction problems.  ... 
doi:10.1007/978-3-642-33765-9_48 fatcat:dnmtthkfuvhrdbxnddwadpd3gy

Effective Melanoma Recognition Using Deep Convolutional Neural Network with Covariance Discriminant Loss

Lei Guo, Gang Xie, Xinying Xu, Jinchang Ren
2020 Sensors  
Aiming at the issues, we propose a melanoma recognition method using deep convolutional neural network with covariance discriminant loss in dermoscopy images.  ...  On the International Symposium on Biomedical Imaging (ISBI) 2018 Skin Lesion Analysis dataset, the two schemes in the proposed method can yield a sensitivity of 0.942 and 0.917, respectively.  ...  The contributions of our work are summarized as follows: (a) (b) (c) (d) (1) We propose a melanoma recognition approach with covariance discriminant (CovD) loss and DCNN, ensuring feature representation  ... 
doi:10.3390/s20205786 pmid:33066123 fatcat:mzavhdh2yjeypacyrufnbamkiy

High-Dimensional Quadratic Discriminant Analysis under Spiked Covariance Model [article]

Houssem Sifaou, Abla Kammoun, Mohamed-Slim Alouini
2020 arXiv   pre-print
Quadratic discriminant analysis (QDA) is a widely used classification technique that generalizes the linear discriminant analysis (LDA) classifier to the case of distinct covariance matrices among classes  ...  For the QDA classifier to yield high classification performance, an accurate estimation of the covariance matrices is required.  ...  In accordance with the covariance matrix model in (1) , it is natural to set t p−r,i = · · · = t p,i = 1/σ 2 i .  ... 
arXiv:2006.14325v1 fatcat:3zjsp5vunfggxjvpeoleum5rdm

High-Dimensional Quadratic Discriminant Analysis under Spiked Covariance Model

Houssem Sifaou, Abla Kammoun, Mohamed-Slim Alouini
2020 IEEE Access  
In accordance with the covariance matrix model in (1) , it is natural to set t p−r,i = · · · = t p,i = 1/σ 2 i .  ...  Moreover, a larger training set is needed for these classifiers to approach the performance of Imp-QDA.  ...  We shall also recall the following formula allowing to compute the variance and covariance of quadratic forms of a multivariate normal distribution.  ... 
doi:10.1109/access.2020.3004812 fatcat:xqcfbo7445cebnkq7gagv6f42q

Multi-class object recognition using boosted linear discriminant analysis combined with masking covariance matrix method

M. Tanigawa
2006 Fourth IEEE International Conference on Computer Vision Systems (ICVS'06)  
We propose a new algorithm, boosted linear discriminant analysis (bLDA), for classification of a non-linear pattern distribution, and masking covariance matrix method (MCM) for robust and fast computation  ...  In MCM, the covariance vectors of training data set are restricted to be locally correlated, by multiplication of covariance mask. bLDA combined with MCM automatically and effectively extract spatially-local  ...  Conclusion We proposed boosted linear discriminant analysis (bLDA) and masking covariance matrix method (MCM). bLDA combined with MCM extracts effective local features and performs high discriminant ability  ... 
doi:10.1109/icvs.2006.44 dblp:conf/icvs/Tanigawa06 fatcat:7secof45wjfktnmdqhkfn3epqa

Pairwise-Covariance Multi-view Discriminant Analysis for Robust Cross-view Human Action Recognition

Hoang-Nhat Tran, Hong Quan Nguyen, Huong Giang Doan, Thanh-Hai Trana, Thi-Lan Le, Hai Vu
2021 IEEE Access  
In this paper, we propose a novel method that leverages successful deep learning-based features for action representation and multi-view analysis to accomplish robust HAR under viewpoint changes.  ...  To this end, we first adopt Multi-view Discriminant Analysis (MvDA).  ...  In [20] , a supervised approach named Multi-view Fisher Discriminant Analysis (MvFDA) was proposed for binary classification problem.  ... 
doi:10.1109/access.2021.3082142 fatcat:pbsa32rzfrc2zekla5rge7wy3u

Set2Model Networks: Learning Discriminatively To Learn Generative Models

Andrey Kuzmin, Alexander Vakhitov, Victor Lempitsky
2017 2017 IEEE International Conference on Computer Vision Workshops (ICCVW)  
We present a new "learning-to-learn"-type approach for small-to-medium sized training sets.  ...  A trained Set2Model network facilitates learning in the cases when no negative examples are available, and whenever the concept being learned is polysemous or represented by noisy training sets.  ...  The approach [20] learns a metric in the image feature domain in order to improve distance-based image classification and shows that the resulting metric generalizes well to the classes unseen during  ... 
doi:10.1109/iccvw.2017.50 dblp:conf/iccvw/KuzminVL17 fatcat:fj6phvolcfcploi2p7lzzx2tka

Learning a perceptual manifold for image set classification

Sriram Kumar, Andreas Savakis
2016 2016 IEEE International Conference on Image Processing (ICIP)  
We demonstrate the efficacy of our approach for image set classification on face and object recognition datasets.  ...  We present a biologically motivated manifold learning framework for image set classification inspired by Independent Component Analysis for Grassmann manifolds.  ...  In image set classification, each set consists of a number of images that belong to the same class and can capture natural variations in the object's appearance, e.g. due to pose changes and varying illumination  ... 
doi:10.1109/icip.2016.7533198 dblp:conf/icip/KumarS16 fatcat:vpe5njldcrchnic5r54b4ecz6i

Representation Learning with Deep Extreme Learning Machines for Efficient Image Set Classification [article]

Muhammad Uzair, Faisal Shafait, Bernard Ghanem, Ajmal Mian
2015 arXiv   pre-print
Efficient and accurate joint representation of a collection of images, that belong to the same class, is a major research challenge for practical image set classification.  ...  We learn the non-linear structure of image sets with Deep Extreme Learning Machines (DELM) that are very efficient and generalize well even on a limited number of training samples.  ...  Image set classification naturally arises in many applications when a given collection of images are known to belong to one class but with unknown identity.  ... 
arXiv:1503.02445v3 fatcat:2x3pddwt3jagldi2asogmgnvlq

Learning Discriminative Metrics via Generative Models and Kernel Learning [article]

Yuan Shi, Yung-Kyun Noh, Fei Sha, Daniel D. Lee
2011 arXiv   pre-print
In this paper, we show how to unify generative and discriminative learning of metrics via a kernel learning framework.  ...  The proposed learning algorithm is also very efficient, achieving order of magnitude speedup in training time compared to previous discriminative baseline methods.  ...  However, M KDE is less efficient than M UNI due to its iterative nature. We plan to explore more efficient and theoretical-sound combination approach in our future work.  ... 
arXiv:1109.3940v1 fatcat:totbvwpmmjbyzpanvhacacsptu

Manifold learning for brain connectivity [article]

Félix Renard, Christian Heinrich, Marine Bouthillon, Maleka Schenck, Francis Schneider, Stéphane Kremer, Sophie Achard
2020 arXiv   pre-print
In this context, our goal is to provide a machine learning technique that exhibits flexibility, gives the investigator grip on the features and covariates, allows visualization and exploration, and yields  ...  The retained approach is dimension reduction in a manifold learning methodology, the originality lying in that one (or several) reduced variables be chosen by the investigator.  ...  A random graph will have a global efficiency close to 1 for each node, and a regular graph will have a global efficiency close to 0 for each node.  ... 
arXiv:2005.00469v1 fatcat:bvc2zj5vibblpecdrbs5egckvi
« Previous Showing results 1 — 15 out of 21,877 results