Isaac Martín de Diego

Methods for the combination of kernel matrices within a support vector framework

The problem of combining different sources of information arises in several situations, for instance, the classification of data with asymmetric similarity matrices or the construction of an optimal classifier from a collection of kernels. Often, each source of information can be expressed as a similarity matrix. In this paper we propose a new class of methods in order to produce, for classification purposes, a single kernel matrix from a collection of kernel (similarity) matrices. Then, the constructed kernel matrix is used to train a Support Vector Machine (SVM). The key ideas within the kernel construction are twofold: the quantification, relative to the classification labels, of the difference of information among the similarities; and the extension of the concept of linear combination of similarity matrices to the concept of functional combination of similarity matrices. The proposed methods have been successfully evaluated and compared with other powerful classifiers and kernel combination techniques on a variety of artificial and real classification problems.

Improving Support Vector Classification via the Combination of Multiple Sources of Information

In this paper we describe several new methods to build a kernel matrix from a collection of kernels. This kernel will be used for classification purposes using Support Vector Machines (SVMs). The key idea is to extend the concept of linear combination of kernels to the concept of functional (matrix) combination of kernels. The functions involved in the combination take advantage of class conditional probabilities and nearest neighbour techniques. The proposed methods have been successfully evaluated on a variety of real data sets against a battery of powerful classifiers and other kernel combination techniques.

Combining Kernel Information for Support Vector Classification

In this paper we describe new methods to built a kernel matrix from a collection of kernels for classification purposes using Support Vector Machines (SVMs). The methods build the combination by quantifying, relative to the classification labels, the difference of information among the kernels. The proposed techniques have been successfully evaluated on a variety of artificial and real data sets.

Support vector machine classifiers for asymmetric proximities

The aim of this paper is to afford classification tasks on asymmetric kernel matrices using Support Vector Machines (SVMs). Ordinary theory for SVMs requires to work with symmetric proximity matrices. In this work we examine the performance of several symmetrization methods in classification tasks. In addition we propose a new method that specifically takes classification labels into account to build the proximity matrix. The performance of the considered method is evaluated on a variety of artificial and real data sets.

From indefinite to positive semi-definite matrices

Similarity based classification methods use positive semi-definite (PSD) similarity matrices. When several data representations (or metrics) are available, they should be combined to build a single similarity matrix. Often the resulting combination is an indefinite matrix and can not be used to train the classifier. In this paper we introduce new methods to build a PSD matrix from an indefinite matrix. The obtained matrices are used as input kernels to train Support Vector Machines (SVMs) for classification tasks. Experimental results on artificial and real data sets are reported.

Face-based recognition systems in the ABC e-gates

Automated border control (ABC) systems have been proposed as the best option to satisfy the actual needs that airport security has to face nowadays and in the next years. These needs are demanded by the rapid flow growth. High efficient biometric recognition systems using face, iris or fingerprint modalities placed inside the gates at airports entry points should cope with the problems caused by this growth. Problems such as, congestion at these gates, delays in the planned arrival schedules or lack of security control, require a fast automated biometric solution. Face modality is present in 2nd generation passports and it is well accepted by travellers. In this work the state of the art for the most important face recognition models that are currently employed in ABC e-Gates is presented. The importance that image quality has in the ABC systems performance and how external factors affect this quality is described.

Subjective Traffic Safety Experts' Knowledge for Driving-Risk Definition

This paper presents a novelty system for the detection of driving-risk situations based on the knowledge acquired from traffic safety experts. A complete methodology to generate a driving-risk reference signal has been developed. A set of driving sessions was executed in a very realistic truck simulator, where several measures and visual information from the vehicle, the driver, and the road were collected. Two kinds of experiments were designed, i.e., controlled driving sessions (where several risk situations were induced) and natural driving sessions (where no risk situations were induced and a natural driving behavior was expected). A group of traffic safety experts from the Royal Automobile Club of Spain was consulted to evaluate the driving risk in each simulated session. The information acquired from the traffic safety experts was used to develop a methodology to combine their evaluations. The risks detected with the proposed methodology were analyzed to determine the most common human factors related with the generation of driving-risk situations.

Face recognition improvement with distortions of images in training set

The results and comparative analysis of two face recognition methods are presented in this article. The Permutation Coding Neural Classifier (PCNC) and Support Vector Machine (SVM) methods were selected. The main idea is to improve the image recognition rate. For this purpose we increase the training set by including distortions of initial images. Different numbers of distortions can increase the training image set and improve the quality of the classifier. The goal is to investigate the influence of the number of distortions on the PCNC recognition rate. Using distortions it is possible to improve the PCNC recognition rate. Sometimes it is possible to decrease the number of errors by 10 times or more. The PCNC with 12 distortions outperforms the results of SVM.

Incremental Learning with Soft-Biometric Features for People Re-Identification in Multi-Camera Environments

In this paper, a solution for the appearance based people re-identification problem in a non-overlapping multicamera surveillance environment is presented. For this purpose, an incremental learning approach and a SVM classifier have been considered. The proposed methods update the appearance model across different camera conditions in three different ways: based on time lapses, on change of camera and on the automatic selection of the most representative samples. In order to test the proposed methods, a complete database was acquired at Barajas international airport (the MUBA proposed database). Further the well known PETS 2006 and PETS 2009 databases were considered. The system has been designed for video surveillance security. The main idea of this system is that, in an initial point, the suspect is manually identified by the user. Then, from that moment, the system is able to identify the selected subject across the different cameras in the surveillance area. The results obtained show the importance of the model update and the huge potential of the incremental learning approach.

Alternatives to parameter selection for kernel methods

In this paper we propose alternative methods to parameter selection techniques in order to build a kernel matrix for classification purposes using Support Vector Machines (SVMs). We describe several methods to build a unique kernel matrix from a collection of kernels built using a wide range of values for the unkown parameters. The proposed techniques have been successfully evaluated on a variety of artificial and real data sets. The new methods outperform the best individual kernel under consideration and they can be used as an alternative to the parameter selection problem in kernel methods.