Hubert Cardot

Cooperation of color pixel classification schemes and color watershed: a study for microscopic images

In this paper, we study the ability of the cooperation of two-color pixel classification schemes (Bayesian and K-means classification) with color watershed. Using color pixel classification alone does not sufficiently accurately extract color regions so we suggest to use a strategy based on three steps: simplification, classification, and color watershed. Color watershed is based on a new aggregation function using local and global criteria. The strategy is performed on microscopic images. Quantitative measures are used to evaluate the resulting segmentations according to a learning set of reference images.

Fusion of methods for keystroke dynamic authentication

In this article, we present three methods for the keystroke dynamic authentication problem. We use in the first method, the average and the standard deviation, in the second the rhythm of striking and in the third, a comparison of the times order. After having presented these methods, we propose to realize a fusion of them. The results obtained indicate good performance of each method alone, as well as a significant improvement of performance with fusion, from 3.43% of EER for the best method alone down-to 1.8% with fusion.

A color object recognition scheme: application to cellular sorting

Abstract. This paper presents a color object recognition scheme which proceeds in three sequential steps: segmentation, features extraction and classification. We mainly focus on the first and the third steps here. A color watershed using global and local criteria is first described. A color contrast value is defined to select the best color space for segmenting color objects. Then, an architecture of binary neural networks is described. Its properties relies on the simplification of the recognition problem, leading to a noticeable increase in the classification rate. We conclude with the abilities of such a recognition scheme and present an automated cell sorting system.

SVM training time reduction using vector quantization

In this paper, we describe a new method for training SVM on large data sets. Vector quantization is applied to reduce a large data set by replacing examples by prototypes. Training time for choosing optimal parameters is greatly reduced. Some experimental results yields to demonstrate that this method can reduce training time by a factor of 100, while preserving classification rate. Moreover this method allows to find a decision function with a low complexity when the training data set includes noisy or error examples.

A neural network architecture for data classification

This article aims at showing an architecture of neural networks designed for the classification of data distributed among a high number of classes. A significant gain in the global classification rate can be obtained by using our architecture. This latter is based on a set of several little neural networks, each one discriminating only two classes. The specialization of each neural network simplifies their structure and improves the classification. Moreover, the learning step automatically determines the number of hidden neurons. The discussion is illustrated by tests on databases from the UCI machine learning database repository. The experimental results show that this architecture can achieve a faster learning, simpler neural networks and an improved performance in classification.

TABU SEARCH MODEL SELECTION FOR SVM

A model selection method based on tabu search is proposed to build support vector machines (binary decision functions) of reduced complexity and efficient generalization. The aim is to build a fast and efficient support vector machines classifier. A criterion is defined to evaluate the decision function quality which blends recognition rate and the complexity of a binary decision functions together. The selection of the simplification level by vector quantization, of a feature subset and of support vector machines hyperparameters are performed by tabu search method to optimize the defined decision function quality criterion in order to find a good sub-optimal model on tractable times.

Estimation of User Specific Parameters in One-class Problems

In this paper, we propose a method to find and use user-dependant parameters to increase the performance of a keystroke dynamic system. These parameters include the security threshold and fusion weights of different classifiers. We have determined a set of global parameters which increase the performance of some keystroke dynamics methods. Our experiments show that parameter personalization greatly increases the performance. The main problem is how to estimate the parameters from only a user training set containing ten login sequences. This problem is a promising way to increase performance in biometric but it is still an open path

Bayesian marker extraction for color watershed in segmenting microscopic images

In this paper we study the ability of the cooperation of Bayesian color pixel classification in extracting seeds for color watershed. Using color pixel classification alone does not extract accurately enough color regions so we suggest to use a strategy based on three steps: simplification, Bayesian classification and color watershed color watershed is based on an aggregation function using local and global criteria. The strategy is performed on microscopic images. Quantitative measures are used to evaluate the resulting segmentations according to a set of reference images.

Automatic system for the classification of cellular categories in cytological images

In this paper, we describe research carried out within the framework of the optimization of an image analyzer dedicated to rapid detection of abnormalities of ploidy in human tumors. The system takes as its input microscopic images of dissociated cells which are to be segmented in order to extract cellular objects, calculate shape and texture measures, and identify each category of cell, by means of two classification methods that are compared and discussed: classification based on the Bayes decision rule and classification using neural networks.

Neural network induction graph for pattern recognition

Abstract This paper presents a novel architecture of neural networks designed for pattern recognition. The concept of induction graphs coupled with a divide-and-conquer strategy defines a neural network induction graph (NNIG). First, the NNIG concept is described and its properties detailed. It is based on a set of several little neural networks, each one discriminating only two classes. The specialization of each neural network simplifies their structure and improves the classification. The principle used to perform the decision of classification on an input pattern is explained. The latter enables to take into account dubious decisions identified by the NNIG. The last section presents experimental results. A significant gain in the global classification rate can be obtained by using an NNIG. The discussion is illustrated by tests on databases from the UCI machine learning database repository. The experimental results show that an NNIG can achieve a better learning, simpler neural networks and an improved performance in classification. A final illustration is presented on a real microscopical imaging problem for the classification of cells in serous cytology.