Search results for "pattern recognition"

A Robust Multi Stage Technique for Image Binarization of Degraded Historical Documents

2017

International audience; Document image binarization is a central problem in many document analysis systems. Indeed, it represents one of the basic challenges, especially in case of historical documents analysis. In this paper, we propose a novel robust multi stage framework that combines different existing document image thresholding methods for the purpose of getting a better binarization result. CLAHE technique is introduced to significantly enhance contrast in some poor images. The proposed method then uses a hybrid algorithm to partition image into foreground and background. A special procedure is finally applied in order to remove small noise and correct characters morphology. Experime…

SHREC'12 Track: 3D mesh segmentation

2012

International audience; 3D mesh segmentation is a fundamental process in many applications such as shape retrieval, compression, deformation, etc. The objective of this track is to evaluate the performance of recent segmentation methods using a ground-truth corpus and an accurate similarity metric. The ground-truth corpus is composed of 28 watertight models, grouped in five classes (animal, furniture, hand, human and bust) and each associated with 4 ground-truth segmentations done by human subjects. 3 research groups have participated to this track, the accuracy of their segmentation algorithms have been evaluated and compared with 4 other state-of-the-art methods.

2020

Although the individuality of whole-body movements has been suspected for years, the scientific proof and systematic investigation that individuals possess unique movement patterns did not manifest until the introduction of the criteria of uniqueness and persistence from the field of forensic science. Applying the criteria of uniqueness and persistence to the individuality of motor learning processes requires complex strategies due to the problem of persistence in the learning processes. One approach is to examine the learning process of different movements. For this purpose, it is necessary to differentiate between two components of movement patterns: the individual-specific component and …

Rôle de l’inflammasome dans les pathologies cardiovasculaires

2011

NOD-like receptors (NLRs) constitute a recently identified family of intracellular pattern recognition receptors which contains more than 20 members in mammals. Some of the NLRs, the NALP subfamily, constituted from 14 members, many of them without actual identified role, form multiproteic complex known as inflammasome, that initiate inflammation by processing inactive pro-caspase-1 to its active form, allowing the cleavage and subsequent activation of pro-IL-1β and pro-IL-18. We review the identified roles of NLRs in pathologies and argue for the role of inflammasome in the development of cardiovascular diseases. The atherogenic cytokines IL-1β and IL-18 are matured in NLRPs inflammasomes.…

Convergence Analysis of Distributed Set-Valued Information Systems

2016

This paper focuses on the convergence of information in distributed systems of agents communicating over a network. The information on which the convergence is sought is not rep- resented by real numbers, as often in the literature, rather by sets. The dynamics of the evolution of information across the net- work is accordingly described by set-valued iterative maps. While the study of convergence of set-valued iterative maps is highly complex in general, this paper focuses on Boolean maps, which are comprised of arbitrary combinations of unions, intersections, and complements of sets. For these important class of systems, we provide tools to study both global and local convergence. A distr…

Learning-based multiresolution transforms with application to image compression

2013

In Harten's framework, multiresolution transforms are defined by predicting finer resolution levels of information from coarser ones using an operator, called prediction operator, and defining details (or wavelet coefficients) that are the difference between the exact and predicted values. In this paper we use tools of statistical learning in order to design a more accurate prediction operator in this framework based on a training sample, resulting in multiresolution decompositions with enhanced sparsity. In the case of images, we incorporate edge detection techniques in the design of the prediction operator in order to avoid Gibbs phenomenon. Numerical tests are presented showing that the …

Predicting perceptual distortion sensitivity with gain control models of LGN

2017

Classification of persimmon fruit origin by near infrared spectrometry and least squares-support vector machines

2014

Abstract The main objective of this work has been the authentication by Fourier transform near infrared (FT-NIR) spectrometry of the origin of persimmon fruits cultivated in different regions of Spain. In order to achieve this goal, 166 persimmon samples from 7 different regions of Spain were analyzed by FT-NIR spectrometry. By splitting the spectral data in training and independent test sets, a classification model was built using least squares support vector machines chemometric technique. Orthogonal signal correction and principal component analysis were performed prior to conduct the classification strategy. The verified model was applied for the prediction of the origin of 50 samples f…

Alternative method for binary shape alignment of non-symmetrical shapes based on minimal enclosing box

2012

Proposed is a novel method based on the minimal enclosing box (MEB) to determine the canonical orientation associated with a three-dimensional binary shape. It is suggested that, when the shape has no clear distinctive features and two or more of the eigenvalues are similar, this method is more suitable than the commonly used method based on principal component analysis (PCA). An experiment is performed with shapes of human livers by measuring the degree on which a prototypical image (atlas) matches to a new shape after alignment by PCA, minimal area projection (MAP), and MEB showing that in this case MEB outperforms the usual PCA-based alignment method and also the MAP method.

Ensemble feature selection with the simple Bayesian classification

2003

Abstract A popular method for creating an accurate classifier from a set of training data is to build several classifiers, and then to combine their predictions. The ensembles of simple Bayesian classifiers have traditionally not been a focus of research. One way to generate an ensemble of accurate and diverse simple Bayesian classifiers is to use different feature subsets generated with the random subspace method. In this case, the ensemble consists of multiple classifiers constructed by randomly selecting feature subsets, that is, classifiers constructed in randomly chosen subspaces. In this paper, we present an algorithm for building ensembles of simple Bayesian classifiers in random sub…