Filters








2,752 Hits in 5.1 sec

Closed-Form Jensen-Renyi Divergence for Mixture of Gaussians and Applications to Group-Wise Shape Registration [chapter]

Fei Wang, Tanveer Syeda-Mahmood, Baba C. Vemuri, David Beymer, Anand Rangarajan
2009 Lecture Notes in Computer Science  
Jensen-Renyi (JR) divergence.  ...  We evaluate a closed-form JR divergence between multiple probabilistic representations for the general case where the mixture models differ in variance and the number of components.  ...  We first demonstrate the robustness and accuracy that can be achieved in registration with JR divergence in comparison to Jensen-Shannon divergence reported in an earlier work.  ... 
doi:10.1007/978-3-642-04268-3_80 fatcat:pw3aruf33fef7jho2hfhjlihmu

Generalized L2-Divergence and Its Application to Shape Alignment [chapter]

Fei Wang, Baba Vemuri, Tanveer Syeda-Mahmood
2009 Lecture Notes in Computer Science  
We derive a closed-form expression for the Generalized-L2 divergence between multiple Gaussian mixtures, which in turn leads to a computationally efficient registration algorithm.  ...  We develop a novel divergence measure which is defined between any arbitrary number of probability distributions based on L2 distance, and we call this new divergence measure "Generalized L2divergence"  ...  They minimize the Jensen-Shannon divergence and the CDF-based Jensen-Shannon divergence respectively between the feature point-sets with respect to non-rigid deformation.  ... 
doi:10.1007/978-3-642-02498-6_19 fatcat:jbylqlfnzfc3zcjj3saarwlifu

AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty [article]

Dan Hendrycks, Norman Mu, Ekin D. Cubuk, Barret Zoph, Justin Gilmer, Balaji Lakshminarayanan
2020 arXiv   pre-print
In this work, we propose a technique to improve the robustness and uncertainty estimates of image classifiers.  ...  We propose AugMix, a data processing technique that is simple to implement, adds limited computational overhead, and helps models withstand unforeseen corruptions.  ...  Variety through randomness, the Jensen-Shannon divergence (JSD) loss, and augmentation mixing confer robustness.  ... 
arXiv:1912.02781v2 fatcat:s5a7xpmjwjayphmrzqexkxzyni

Estimating equilibrium ensemble averages using multiple time slices from driven nonequilibrium processes: Theory and application to free energies, moments, and thermodynamic length in single-molecule pulling experiments

David D. L. Minh, John D. Chodera
2011 Journal of Chemical Physics  
Our expression may be combined with estimators of path-ensemble averages, including existing optimal estimators that use data collected by unidirectional and bidirectional protocols.  ...  Compared to estimators that only use individual time slices, our multiple time-slice estimators yield substantially smoother estimates and achieve lower variance for higher-order moments.  ...  Estimates of the Jensen-Shannon divergence.  ... 
doi:10.1063/1.3516517 pmid:21241084 fatcat:4m2uxr72mrbutnwojx752if3jy

Precise Segmentation of Multiple Organs in CT Volumes Using Learning-Based Approach and Information Theory [chapter]

Chao Lu, Yefeng Zheng, Neil Birkbeck, Jingdan Zhang, Timo Kohlberger, Christian Tietjen, Thomas Boettger, James S. Duncan, S. Kevin Zhou
2012 Lecture Notes in Computer Science  
The incorporation of the Jensen-Shannon divergence further drives the mesh to the best fit of the image, thus improves the segmentation performance.  ...  First, marginal space learning (MSL) is applied to efficiently and effectively localize the multiple organs in the largely diverse CT volumes.  ...  Jensen-Shannon (JS) divergence, first introduced in [10], serves as a measure of cohesion between multiple probability distributions.  ... 
doi:10.1007/978-3-642-33418-4_57 fatcat:3f4zbqdysjgu3gxjmprnkfripm

Simultaneous Nonrigid Registration of Multiple Point Sets and Atlas Construction

F. Wang, B.C. Vemuri, A. Rangarajan, S.J. Eisenschenk
2008 IEEE Transactions on Pattern Analysis and Machine Intelligence  
This algorithm avoids the correspondence problem by minimizing the Jensen-Shannon (JS) divergence between the point sets represented as finite mixtures of Gaussian densities.  ...  In this paper, we propose a novel and robust algorithm that is capable of simultaneously computing the mean shape-represented by a probability density functionfrom multiple unlabeled point-sets (represented  ...  ., N }, each with a weight π a in the Jensen-Shannon divergence.  ... 
doi:10.1109/tpami.2007.70829 pmid:18787248 pmcid:PMC2921641 fatcat:hfuhclej7rbjfi5hwcvuknaube

Simultaneous Nonrigid Registration of Multiple Point Sets and Atlas Construction [chapter]

Fei Wang, Baba C. Vemuri, Anand Rangarajan, Ilona M. Schmalfuss, Stephan J. Eisenschenk
2006 Lecture Notes in Computer Science  
This algorithm avoids the correspondence problem by minimizing the Jensen-Shannon (JS) divergence between the point sets represented as finite mixtures of Gaussian densities.  ...  In this paper, we propose a novel and robust algorithm that is capable of simultaneously computing the mean shape-represented by a probability density functionfrom multiple unlabeled point-sets (represented  ...  ., N }, each with a weight π a in the Jensen-Shannon divergence.  ... 
doi:10.1007/11744078_43 fatcat:ruuatkpwafckvo7b52evlfn5le

Gaussian-weighted Jensen–Shannon divergence as a robust fitness function for multi-model fitting

Kai Zhou, Karthik Mahesh Varadarajan, Michael Zillich, Markus Vincze
2013 Machine Vision and Applications  
In this paper, we present a novel model evaluation function based on Gaussian-weighted Jensen-Shannon divergence, and integrate into a particle swarm optimization (PSO) framework using ring topology.  ...  Superior performance in terms of processing time and robustness to inlier noise is also demonstrated with respect to state of the art methods.  ...  Gaussian-weighted Jensen-Shannon divergence This section describes the details of the proposed evaluation function using Gaussian-weighted Jensen-Shannon divergence [18] .  ... 
doi:10.1007/s00138-013-0513-1 fatcat:splhbaeyy5eypioy5qsxc2idmy

A New Information-Theoretic Measure to Control the Robustness-Sensitivity Trade-Off for DMFFD Point-Set Registration [chapter]

Nicholas J. Tustison, Suyash P. Awate, Gang Song, Tessa S. Cook, James C. Gee
2009 Lecture Notes in Computer Science  
Wang et al. generalize the Kullback-Leibler (KL) divergence measure to the Jensen-Shannon divergence measure for the construction of unbiased atlases [7] using a thin-plate spline transformation model.  ...  The proposed measure is a generalization of the Jenson-Shannon divergence known as the Jensen-Havrda-Charvat-Tsallis divergence.  ...  Discussion and Conclusions We have described a technique for point-set correspondence based on a generalization of the Jensen-Shannon divergence.  ... 
doi:10.1007/978-3-642-02498-6_18 fatcat:7qvj2a3ngffejhvjwinldp2vmu

Jensen-Shannon Divergence for Non-destructive Incipient Crack Detection and Estimation

Xiaoxia Zhang, Claude Delpha, Demba Diallo
2020 IEEE Access  
Jensen-Shannon (JSD) divergence is then proposed for the crack detection. Thanks to a theoretical derivation, the fault severity estimation is obtained.  ...  In this paper, we highlight the limitation of classical techniques and address this problem using a methodology based on wavelet transform and Jensen-Shannon divergence in the framework of Noisy Independent  ...  We propose in this work, an estimation model depending on the Jensen-Shannon Divergence. This model is derived taking into account our particular feature space.  ... 
doi:10.1109/access.2020.3004658 fatcat:mspdm4w7wveqjdejc6avqdjaiy

Computer Analysis of Images and Patterns

Pedro Real
2013 Journal of Mathematical Imaging and Vision  
The next section studies representation and estimation of shape from different mathematical points of view. P. Real ( ) Dpto.  ...  "Tuning of Adaptive Weight Depth Map Generation Algorithms" by Acosta et al. shows a systematic statistical approach for the parameter setting problem in disparity map estimation within the context of  ...  The last paper of the section, "Graph Kernels from the Jensen-Shannon Divergence" by Bai et al., shows how to construct Jensen-Shannon kernels for graph datasets using the von-Neumann entropy and Shannon  ... 
doi:10.1007/s10851-013-0451-6 fatcat:birxsnfbhbei3az57ajaydsr5i

Evaluating Bregman Divergences for Probability Learning from Crowd [article]

F. A. Mena, R. Ñanculef
2019 arXiv   pre-print
We focus on the Bregman divergences framework to used as objective function to minimize.  ...  Here we present differents models that adapts having probability distribution as target to train a machine learning model.  ...  For example making KL symmetric (Jensen-Shannon divergence) improve some results on the test set.  ... 
arXiv:1901.10653v1 fatcat:554otnydlba5tevtewcoyvn6na

Total Jensen divergences: Definition, Properties and k-Means++ Clustering [article]

Frank Nielsen, Richard Nock
2013 arXiv   pre-print
We then proceed by defining the total Jensen centroids as average distortion minimizers, and study their robustness performance to outliers.  ...  We present a novel class of divergences induced by a smooth convex function called total Jensen divergences.  ...  Total Bregman divergence, skew Jensen divergence, Jensen-Shannon divergence, Burbea-Rao divergence, centroid robustness, Stolarsky mean, clustering, k-means++. I.  ... 
arXiv:1309.7109v1 fatcat:46zvueh3hbcdleuwwjjkrhxrya

Attribute vector guided groupwise registration

Qian Wang, Guorong Wu, Pew-Thian Yap, Dinggang Shen
2010 NeuroImage  
For the purpose of registration, Jensen-Shannon (JS) divergence is used to measure the PDF dissimilarity of each attribute at corresponding locations of different individual images.  ...  By minimizing the overall JS divergence in the whole image space and estimating the deformation field of each image simultaneously, we can eventually register all images and build an unbiased atlas.  ...  Jensen-Shannon Divergence and PDF Stack Divergence Jensen-Shannon (JS) divergence [11] is an information-theoretic quantity for measuring the dissimilarity or the distance of two probability distributions  ... 
doi:10.1016/j.neuroimage.2010.01.040 pmid:20097291 pmcid:PMC2839051 fatcat:ohrilhwbezhrrauusb4a72uoja

Attribute Vector Guided Groupwise Registration [chapter]

Qian Wang, Pew-Thian Yap, Guorong Wu, Dinggang Shen
2009 Lecture Notes in Computer Science  
For the purpose of registration, Jensen-Shannon (JS) divergence is used to measure the PDF dissimilarity of each attribute at corresponding locations of different individual images.  ...  By minimizing the overall JS divergence in the whole image space and estimating the deformation field of each image simultaneously, we can eventually register all images and build an unbiased atlas.  ...  Jensen-Shannon Divergence and PDF Stack Divergence Jensen-Shannon (JS) divergence [11] is an information-theoretic quantity for measuring the dissimilarity or the distance of two probability distributions  ... 
doi:10.1007/978-3-642-04268-3_81 fatcat:s34czletfneu5billf4bkfl2y4
« Previous Showing results 1 — 15 out of 2,752 results