Filters








165,106 Hits in 7.2 sec

Fast Randomized Algorithms for Robust Estimation of Location [chapter]

Vladimir Estivill-Castro, Michael E. Houle
2001 Lecture Notes in Computer Science  
In this paper, we propose O(Dn p n)-time randomized algorithms for computing robust estimators of location, where n is the size of the data set, and D is the dimension.  ...  Most estimators of location exhibiting useful robustness properties require at least quadratic time to compute, far too slow for large data mining applications.  ...  In this paper, we propose a new family of non-deterministic (yet robust) estimators of location that rely on fast randomization techniques for their computation.  ... 
doi:10.1007/3-540-45244-3_7 fatcat:mkhxhsy2dvdtjowqdf6hxnzsui

The DetS and DetMM estimators for multivariate location and scatter

Mia Hubert, Peter Rousseeuw, Dina Vanpaemel, Tim Verdonck
2015 Computational Statistics & Data Analysis  
New deterministic robust estimators of multivariate location and scatter are presented.  ...  Their computation time is much lower than FastS and FastMM, which allows to compute the estimators for a range of breakdown values.  ...  Acknowledgments We thank Clayton Scott for providing the flow cytometry data set.  ... 
doi:10.1016/j.csda.2014.07.013 fatcat:bqb2xvor4rdhtjcle67bwvp2ci

A Fast Algorithm for the Minimum Covariance Determinant Estimator

Peter J. Rousseeuw, Katrien Van Driessen
1999 Technometrics  
The minimum covariance determinant (MCD) method of Rousseeuw (1984) is a highly robust estimator of multivariate location and scatter.  ...  To deal with such problems we have developed a new algorithm for the MCD, called FAST-MCD.  ...  In conclusion, we personally prefer the FAST-MCD algorithm because it is both robust and fast, even for large n.  ... 
doi:10.1080/00401706.1999.10485670 fatcat:7gfxkaahbjdblawybd7m6sbppi

Fast Adaptive Scene Sampling for Single-Photon 3D Lidar Images

Abderrahim Halimi, Philippe Ciuciu, Aongus McCarthy, Stephen McLaughlin, Gerald S. Buller
2019 2019 IEEE 8th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP)  
Based on data statistics, the approach starts by achieving a robust estimation of the depth image.  ...  This paper presents a new approach for adaptive scene sampling allowing for faster acquisition when compared to classical uniform sampling or random sampling strategies.  ...  Note finally that other robust estimation strategies can also be considered provided that the estimation is fast and robust [15] .  ... 
doi:10.1109/camsap45676.2019.9022519 dblp:conf/camsap/HalimiCMMB19 fatcat:ht3y7i4pkrfstdoumzvv7h5dxq

Long-term correlation tracking

Chao Ma, Xiaokang Yang, Chongyang Zhang, Ming-Hsuan Yang
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
Extensive experimental results on large-scale benchmark datasets show that the proposed algorithm performs favorably against state-of-the-art methods in terms of efficiency, accuracy, and robustness.  ...  In addition, we train an online random fern classifier to re-detect objects in case of tracking failure.  ...  For each s ∈ S, we extract an image patch J s of size sP × sQ centered around the estimated location.  ... 
doi:10.1109/cvpr.2015.7299177 dblp:conf/cvpr/MaYZY15 fatcat:45hg4wqx75esrnolt75um7yjs4

Practical edge finding with a robust estimator

Fleck
1994 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition CVPR-94  
This paper presents a new algorithm for locating the boundaries of textured regions (both step changes and outliers) using a robust estimator.  ...  The algorithm is made as fast as a Marr-Hildreth edge nder by restructuring the estimator as a series of 2D image operations, using new multi-ring order statistic operators, and running most of the estimator  ...  Acknowledgements This research was carried out at the Department of Computer Science, University of Iowa, supported by a GE Foundation Faculty Fellowship Grant and by National Science Foundation grants  ... 
doi:10.1109/cvpr.1994.323785 dblp:conf/cvpr/Fleck94 fatcat:i7ggztddifhtbcmfm7r6el7m3y

Computing LTS Regression for Large Data Sets

PETER J. ROUSSEEUW, KATRIEN VAN DRIESSEN
2006 Data mining and knowledge discovery  
It turned out that the computation time of existing LTS algorithms grew too fast with the size of the data set, precluding their use for data mining.  ...  For small data sets FAST-LTS typically nds the exact LTS, whereas for larger data sets it gives more accurate results than existing algorithms for LTS and is faster by orders of magnitude.  ...  The latter will be based on the minimum covariance determinant (MCD) method of (Rousseeuw 1984 (Rousseeuw , 1985 which provides a robust estimate of multivariate location and a robust scatter matrix.  ... 
doi:10.1007/s10618-005-0024-4 fatcat:h54tn7obtjaopoikpntw3o2gga

High breakdown estimation methods for Phase I multivariate control charts

Willis A. Jensen, Jeffrey B. Birch, William H. Woodall
2007 Quality and Reliability Engineering International  
Based on previous studies, it is not clear which of these two estimation methods is best for control chart applications.  ...  The goal of Phase I monitoring of multivariate data is to identify multivariate outliers and step changes so that the estimated control limits are sufficiently accurate for Phase II monitoring.  ...  Because it is not drawing random samples of points, the FAST-MCD algorithm does not have the repeatability issues that are present in the subsampling algorithm.  ... 
doi:10.1002/qre.837 fatcat:aewfssqd2bdmxff6bbkyyc7n3m

Facial Landmark Localization Using Robust Relationship Priors and Approximative Gibbs Sampling [chapter]

Karsten Vogt, Oliver Müller, Jörn Ostermann
2015 Lecture Notes in Computer Science  
Compared with competing methods, our algorithm does not require any prior knowledge or initial guess about the location, scale or pose of the face.  ...  The exceptional robustness of our method is realized by utilizing a L1-loss function and via our new robust shape model based on pairwise topological constraints.  ...  For these landmarks we will therefore select K uniformly distributed bases at random, each inducing an independent estimate of their location.  ... 
doi:10.1007/978-3-319-27863-6_34 fatcat:ez4iujk3lzdfhm73vdf2ab6mr4

Video Tracking Algorithm Based on Kalman Filter and Online Random Forest

Lijun Xue, Lili Wang
2017 International Journal of Multimedia and Ubiquitous Engineering  
For the case of transient target loss, the two methods cooperate with each other to accurately locate the tracking target.  ...  In view of the problem of Kalman filter algorithm, this paper introduces the random forest learning algorithm in the process of Kalman filter tracking.  ...  Acknowledgments This work was financially supported by the National Natural Science Foundation of China (No. 41374129).  ... 
doi:10.14257/ijmue.2017.12.3.05 fatcat:4ryotrkudbf2layp5b5vqjfixq

Fast Computation of Robust Subspace Estimators [article]

Holger Cevallos-Valdiviezo, Stefan Van Aelst
2019 arXiv   pre-print
Alternative algorithms for the robust subspace estimators are proposed that are better suited to compute the solution for high-dimensional problems.  ...  Second, to reduce computation time even further five robust deterministic values are proposed to initialize the algorithms instead of using random starting values.  ...  In Maronna's algorithm 50 random initial orthogonal matrices B q are generated while the initial estimate for the location m is the coordinatewise median of the data.  ... 
arXiv:1803.10290v2 fatcat:t466y36f5zc5fa53jqpr3t5xrq

Minimum covariance determinant and extensions

Mia Hubert, Michiel Debruyne, Peter J. Rousseeuw
2017 Wiley Interdisciplinary Reviews: Computational Statistics  
The Minimum Covariance Determinant (MCD) method is a highly robust estimator of multivariate location and scatter, for which a fast algorithm is available.  ...  The first one is a fast deterministic algorithm which inherits the robustness of the MCD while being almost affine equivariant.  ...  A robust bootstrap for the MCD is proposed in 63 and a fast cross-validation algorithm in 64 . Computation of the MCD for data with missing values is explored in [65] [66] [67] .  ... 
doi:10.1002/wics.1421 fatcat:3hg6hbbsyzbyvojdjig25enusq

Fast Algorithms for Computing High Breakdown Covariance Matrices with Missing Data [chapter]

S. Copt, M.-P. Victoria-Feser
2004 Theory and Applications of Recent Robust Methods  
In the case of robust estimation of multivariate location and scatter, robust covariances have been first investigated by Maronna (1976) .  ...  The estimator of Σ is AΓA T with Γ = diag ¡ σ (Z j ) 2 ¢¯j =1,...,p . A location estimator for µ is given by Aν with ν = (m (Z j ))| j=1,...,p , m() being a (robust) mean function.  ... 
doi:10.1007/978-3-0348-7958-3_7 fatcat:bxwqfkrvlrdgjpywzkfk2tq7ka

Variable Step Size LMS Algorithm using Squared Error and Autocorrelation of Error

Hong Chae Woo
2012 Procedia Engineering  
A variety of different approaches in the variable step adjustment algorithm of the LMS were researched to achieve fast convergence and robustness, but the complexity of the variable step algorithm was  ...  A variable step size LMS algorithm using squared error and autocorrelation of error is proposed to achieve fast convergence and robustness under reasonable complexity.  ...  In many VSS approaches researched, computation complexity is increased for fast convergence and some of VSS algorithms are designed to get robustness.  ... 
doi:10.1016/j.proeng.2012.07.141 fatcat:zkjmuxywfrgchopqf7gvthflaq

Fast motion estimation algorithm using spatial correlation of motion field and hierarchical search

Byung C. Song, Kyoung Won Lim, Jong Beom Ra, Naohisa Ohta
1996 Digital Compression Technologies and Systems for Video Communications  
Our scheme consists of the higher level search that uses spatial correlation for continuous motion and adopts hierarchical structure for random or complex motion, and the lower level search for the final  ...  A new block matching algorithm specially proper for a large search area, is proposed. The algorithm uses spatial correlation of motion field and hierarchical search.  ...  This paper proposes a new fast hierarchical motion estimation algorithm which is especially suitable for a large search area.  ... 
doi:10.1117/12.251300 fatcat:h6xwsiex7fdrxckshmuzuqvjzy
« Previous Showing results 1 — 15 out of 165,106 results