47,537 Hits in 6.1 sec

Automatic optimization of outlier detection ensembles using a limited number of outlier examples

Niko Reunanen, Tomi Räty, Timo Lintonen
2020 International Journal of Data Science and Analytics  
Obtaining examples of outliers can be prohibitively challenging, and the outlier examples should be used efficiently.  ...  We propose an approach for optimizing outlier detection ensembles using a limited number of outlier examples.  ...  A detector is mathematically defined as a function g j (x i ) = s i j , which returns nonnegative real value (outlier score).  ... 
doi:10.1007/s41060-020-00222-4 fatcat:ng5esoj47zch3aswm252wcbykq

Image Stitching using Harris and RANSAC

Rupali Chandratre, V. A Chakkarwar
2014 International Journal of Computer Applications  
RANSAC method is used to choose the closest match between the two images by separating inliers and outliers.  ...  For the Image stitching using the inliers, homography matrix is used which requires least 8 feature points .  ...  of a Image Stitching should combine up.  ... 
doi:10.5120/15706-4567 fatcat:iy4vscl4lbd7te57gn4uj35ogy

Automated Parameters for Troubled-Cell Indicators Using Outlier Detection

M. J. Vuik, J. K. Ryan
2016 SIAM Journal on Scientific Computing  
This is done using Tukey's boxplot approach to detect which coefficients in a vector are straying far beyond others (Tukey, 1977).  ...  This parameter is used in a threshold which decides whether or not to detect an element as a troubled cell. Until now, these parameters could not be chosen automatically.  ...  In particular the authors are grateful to the reviewers for their remarks, which helped us to greatly improve this paper.  ... 
doi:10.1137/15m1018393 fatcat:yrpvwrbdnfhh7cc6hh4f7xtnry

A Method for Locating Digital Evidences with Outlier Detection Using Support Vector Machine

Zaiqiang Liu, Dongdai Lin, Fengdeng Guo
2008 International Journal of Network Security  
In this paper, we introduce a two-tier method to automate the process of locating the digital evidence, which first employ a one-class Support Vector Machine (SVM) outlier detector to filter out insignificant  ...  further analyze the output of the outlier detector to improve the accuracy of investigation.  ...  the Outlier Detector.  ... 
dblp:journals/ijnsec/LiuLG08 fatcat:buuexkwcvzhg5pjx5hifwmwxeu

ECOD: Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions [article]

Zheng Li, Yue Zhao, Xiyang Hu, Nicola Botta, Cezar Ionescu, George H. Chen
2022 arXiv   pre-print
To address these issues, we present a simple yet effective algorithm called ECOD (Empirical-Cumulative-distribution-based Outlier Detection), which is inspired by the fact that outliers are often the "  ...  Our contributions are as follows: (1) we propose a novel outlier detection method called ECOD, which is both parameter-free and easy to interpret; (2) we perform extensive experiments on 30 benchmark datasets  ...  ( §4.2) 2) How effective is ECOD, in comparison to state-of-theart (SOTA) outlier detectors? ( §4.3.1) 3) Under which conditions, the performance of ECOD may degrade?  ... 
arXiv:2201.00382v2 fatcat:7n27b754ongf7jjwamchvs5zoi

Open Set RF Fingerprinting using Generative Outlier Augmentation [article]

Samurdhi Karunaratne, Samer Hanna, Danijela Cabric
2021 arXiv   pre-print
Since acquiring new transmitters to use as known transmitters is highly expensive, we propose to use generative deep learning methods to emulate unauthorized signal samples for the augmentation of training  ...  They argued that if D(z) is an outlier, the logits of the classifier prediction C A (D(z)) i , should be low for all classes i.  ...  Outlier detector architecture and evaluation In [8] , we explored several neural network architectures that could be used for a problem like Eq. 2 such as DClass and OvA.  ... 
arXiv:2108.13099v1 fatcat:ireps6rggrhjpf3j4vwrkoutsi

Patient dosimetry for total body irradiation using single-use MOSFET detectors

Tina Marie Briere, Ramesh Tailor, Naresh Tolani, Karl Prado, Richard Lane, Shiao Woo, Chul Ha, Michael T. Gillin, A. Sam Beddar
2008 Journal of Applied Clinical Medical Physics  
Our results suggest that these detectors could be used for TBI quality assurance monitoring, although TLDs should remain the standard when critical dose measurements are performed.  ...  If OneDose detectors are to be used for TBI, the use of more than one at each location is strongly recommended.  ...  ACKNOWLEDGMENTS The authors are grateful to Sicel Technologies and MedTec for supplying the reader and detectors used in this study.  ... 
doi:10.1120/jacmp.v9i4.2787 pmid:19020482 pmcid:PMC5722362 fatcat:36hnnwy2vrg6tcombpioz25ngi

Outlier identification in stereo correspondences using quadrics

R. DenHollander, A. Hanjalic
2005 Procedings of the British Machine Vision Conference 2005  
Due to the presence of outliers in these correspondences, robust estimation methods need to be used for the computation of the fundamental matrix.  ...  Using local image correspondences, the fundamental matrix describing the epipolar geometry can be estimated.  ...  In practice, the points in the images are derived from interest point detectors and local image neighborhoods are used for matching.  ... 
doi:10.5244/c.19.80 dblp:conf/bmvc/HollanderH05 fatcat:qembeiq275gixiwpqlbq6s3vcu

Revisit network anomaly ranking in datacenter network using re-ranking

Shaohan Huango, Carol Fung, Kui Wang, Yaqi Yang, Zhongzhi Luan, Depei Qian
2015 2015 IEEE 4th International Conference on Cloud Networking (CloudNet)  
We introduce several new features into the re-ranking model to capture extra information about outliers.  ...  There are many network anomaly detection systems being used to identify significant anomalies in datacenter networks.  ...  We use the output of the unsupervised detector, which is a sequence of sorted data points and real sequence result as input for the regression model.  ... 
doi:10.1109/cloudnet.2015.7335302 dblp:conf/cloudnet/HuangoFWYLQ15 fatcat:hupubrzqrnbpxf5csxmmh6rhyi

Elimination Of Clicks From Archive Speech Signals Using Sparse Autoregressive Modeling

Marcin Ciolek, M. Niedzwiecki
2012 Zenodo  
Five outlier detectors are operated simultaneously.  ...  The fifth detector is based on the classical AR model which incorporates only the formant coefficients. The detection alarm is switched on if all five detectors indicate the presence of the outlier.  ... 
doi:10.5281/zenodo.52293 fatcat:mcb4ku2y7bhyfbkqyc47o534ai

An Isolation-based Distributed Outlier Detection Framework using Nearest Neighbor Ensembles for Wireless Sensor Networks

Zhong-Min Wang, Guo-Hao Song, Cong Gao
2019 IEEE Access  
We propose an isolation-based distributed outlier detection framework using nearest-neighbor ensembles (iNNE) to effectively detect outliers in wireless sensor networks.  ...  In addition, we introduce a sliding window to update local detectors, which enables the adaption of dynamic changes in the environment.  ...  LD is the local detector. And buffer stores normal measurements which are used to retrain the local detector.  ... 
doi:10.1109/access.2019.2929581 fatcat:vuys2bvcxjadzea3voyr2wg67u

Color edge detection using vector order statistics

P.E. Trahanias, A.N. Venetsanopoulos
1993 IEEE Transactions on Image Processing  
In the computation of PE. a value of T = 1 (edge height) has been used which is in accordance with the noise level assumed (unit covariance matrix-I,,).  ...  R = \I$") -X(')I1 (2) VR expresses in a quantitative way the deviation of the vector outlier in the highest rank from the vector median in W.  ... 
doi:10.1109/83.217230 pmid:18296214 fatcat:7jzoswcygfcspn3fhuojhca5zy

Using Principal Component Analysis to Solve a Class Imbalance Problem in Traffic Incident Detection

Changjiang Zheng, Shuyan Chen, Wei Wang, Jian Lu
2013 Mathematical Problems in Engineering  
Using principal component analysis (PCA), a one-class classifier for incident detection is constructed from the major and minor principal components of normal instances.  ...  We always used the same values for 𝛼 1 and 𝛼 2 in ( 7 ), as we did not know in advance which type of outliers we should pay more attention to.  ...  [20] proposed that the minor components, ∑ 𝑝 𝑖=𝑝−𝑟+1 (𝑦 2 𝑖 /𝜆 𝑖 ), should be used to detect observations that do not conform to the normal correlation structure.  ... 
doi:10.1155/2013/524861 fatcat:edzvhcqg3jaddk4dxgwjocqhei

Super-resolution of faces using the epipolar constraint

R. Den Hollander, D.-J. De Lange, K. Schutte
2007 Procedings of the British Machine Vision Conference 2007  
First, a face detector is used to find faces in a video frame, after which an optical flow algorithm is applied to track feature points on the faces.  ...  An iterative backprojection method is used for acquiring the super-resolution images.  ...  SR algorithm -per frame The SR method we use is from [10] , which computes the difference between the observed LR frame I LR and the simulated LR frame from the (warped) SR frame I SR .  ... 
doi:10.5244/c.21.42 dblp:conf/bmvc/HollanderLS07 fatcat:l72yd2vfyvavfkvuwv32muux4m

KText: Arbitrary shape text detection using modified K‐Means

Zhuo Qi, Wenyi Chen, Xiaofei Sun, Wangqian Sun, Hui Yang
2021 IET Computer Vision  
Based on that, corresponding text detector is presented, named the Text Detector, based on modified K-Means (KText), which can generate the bounding boundary of word-level texts with an arbitrary shape  ...  Nevertheless, previous methods that grouped characters by learning the relation of adjacent characters or used the heuristic clustering method with a handcrafted feature are unsuitable for dense, curved  ...  WeText first pretrains a light supervised model using some character-level annotations. Then, it applies it to acquire more character samples, which can be used to train a robust character detector.  ... 
doi:10.1049/cvi2.12052 fatcat:mtoyf4eh7baelb2bleoyjtkozq
« Previous Showing results 1 — 15 out of 47,537 results