Dijana Petrovska-Delacrétaz

Learn More
This paper presents an overview of a state-of-the-art text-independent speaker verification system. First, an introduction proposes a modular scheme of the training and test phases of a speaker verification system. Then, the most commonly speech parameterization used in speaker verification, namely, cepstral analysis, is detailed. Gaussian mixture modeling,(More)
With the increasing use of biometrics, more and more concerns are being raised about the privacy of the personal biometric data. Conventional biometric systems store biometric templates in a database. This may lead to the possibility of tracking personal information stored in one database by getting access to another database through cross-database(More)
This article presents an overview of the POLYCOST database dedicated to speaker recognition applications over the telephone network. The main characteristics of this database are: large mixed speech corpus size (> 100 speakers), English spoken by foreigners, mainly digits with some free speech, collected through international telephone lines, and more than(More)
This document describes the acquisition protocols of MyIDea, a new large and realistic multimodal biometric database designed to conduct research experiments in Identity Verification (IV). The key points of MyIDea are threefold: (1) it is strongly multimodal; (2) it implements realistic scenarios in an open-set framework; (3) it uses sensors of different(More)
This paper evaluates the performance of the twelve primary systems submitted to the evaluation on speaker verification in the context of a mobile environment using the MOBIO database. The mobile environment provides a challenging and realistic test-bed for current state-of-the-art speaker verification techniques. Results in terms of equal error rate (EER),(More)
In this article we address the issue of using the Support Vector Learning technique in combination with the currently well performing Gaussian Mixture Models (GMM) for speaker verification experiments. Support Vector Machines (SVM) is a new and very promising technique in statistical learning theory. Recently this technique produced very interesting results(More)
Biometric traits are permanently associated with a user. Though this is an advantage from identity verification point of view, if such biometric data is compromised, it cannot be replaced by a new one and becomes unusable in the system. This limitation can be overcome by combining biometrics with cryptographic techniques to induce revocability in biometric(More)