Bimodal Biometric Person Identification System Under Perturbations
Multibiometric person identification systems play a crucial role in environments where security must be ensured. However, building such systems must jointly encompass a good compromise between computational costs and overall performance. These systems mus
- PDF / 1,189,687 Bytes
- 14 Pages / 430 x 660 pts Page_size
- 61 Downloads / 197 Views
Pontificia Universidad Cat´ olica de Chile Av. Vicu˜ na Mackenna 4860(143), Santiago, Chile [email protected], [email protected] 2 Mathematical Image Analysis Group Faculty of Mathematics and Computer Science Saarland University, Bldg. E11, 66041 Saarbr¨ ucken, Germany [email protected]
Abstract. Multibiometric person identification systems play a crucial role in environments where security must be ensured. However, building such systems must jointly encompass a good compromise between computational costs and overall performance. These systems must also be robust against inherent or potential noise on the data-acquisition machinery. In this respect, we proposed a bimodal identification system that combines two inexpensive and widely accepted biometric traits, namely face and voice information. We use a probabilistic fusion scheme at the matching score level, which linearly weights the classification probabilities of each person-class from both face and voice classifiers. The system is tested under two scenarios: a database composed of perturbation-free faces and voices (ideal case), and a database perturbed with variable Gaussian noise, salt-and-pepper noise and occlusions. Moreover, we develop a simple rule to automatically determine the weight parameter between the classifiers via the empirical evidence obtained from the learning stage and the noise level. The fused recognition systems exceeds in all cases the performance of the face and voice classifiers alone. Keywords: Biometrics, multimodal, identificacion, face, voice, probabilistic fusion, Gaussian noise, salt-and-pepper noise, occlusions.
1
Introduction
Human beings possess a highly developed ability for recognising certain physiological or behavioral characteristics of different persons, particularly under high levels of variability and noise. Designing automatic systems with such capabilities comprises a very complex task with several limitations. Fortunately, in the last few years a large amount of research has been conducted in this direction. Particularly, biometric systems aim at recognising a person based on a set of intrinsic characteristics that the individual possesses. There exist many attributes that can be utilised to build an identification system depending on the application domain [1,2]. The process of combining information from multiple biometric D. Mery and L. Rueda (Eds.): PSIVT 2007, LNCS 4872, pp. 114–127, 2007. c Springer-Verlag Berlin Heidelberg 2007
Bimodal Biometric Person Identification System
115
traits is known as biometric fusion or multimodal biometrics [3]. Multibiometric systems are more robust since they rely on different pieces of evidence before taking a decision. Fusion could be carried out at three different levels: (a) fusion at the feature extraction level, (b) fusion at the matching score level, and (c) fusion at the decision level [4]. Over the last fifteen years several multimodal schemes have been proposed for person identification [5,6,7]. It is known that the face and voice biometrics have lower performance comp
Data Loading...
