A multi-modal approach for high-dimensional feature recognition
- PDF / 680,895 Bytes
- 8 Pages / 595.276 x 790.866 pts Page_size
- 1 Downloads / 183 Views
O R I G I N A L A RT I C L E
A multi-modal approach for high-dimensional feature recognition Kushan Ahmadian · Marina Gavrilova
Published online: 30 October 2012 © Springer-Verlag 2012
Abstract Over the past few decades, biometric recognition firmly established itself as one of the areas of tremendous potential to make scientific discovery and to advance stateof-the- art research in security domain. Hardly, there is a single area of IT left untouched by increased vulnerabilities, on-line scams, e-fraud, illegal activities, and event pranks in virtual worlds. In parallel with biometric development, which went from focus on single biometric recognition methods to multi-modal information fusion, another rising area of research is virtual world’s security and avatar recognition. This article explores links between multi-biometric system architecture and virtual worlds face recognition, and proposes methodology which can be of benefit for both applications. Keywords Neural networks · Face recognition · Biometric · Security · Avatars · Virtual worlds
1 Introduction and motivation Over the past 10 years or so, primarily in response to growing security threats and financial fraud, it has become necessary to be able to accurately authenticate identity of human beings using biometrics. Using modern technologies, design, and implementing secure access control systems based on physical or behavioral traits are becoming more and more crucial. Within the process of recognition/identification, it K. Ahmadian · M. Gavrilova () Department of Computer Science, University of Calgary, Calgary, Canada e-mail: [email protected] K. Ahmadian e-mail: [email protected]
usually happens that a set of features are extracted from the input patterns/images of the user and afterward compared to a set of features stored in the database. Recognition of a system user based on fingerprints, iris, face, voice, gait, or typing pattern are common in many commercial or personal applications. Recently, such technologies made their way into the virtual world, where biometric approaches and multi-model schemes are successfully applied to virtual entity recognition [11, 15]. As a very important biometric feature, facial biometrics plays a significant role in user authentication. It is usually comprised of a large set of high-dimensional vectors representing topological, color, or texture information, which makes it a hard biometric pattern to learn [8, 12]. Many of the earlier face recognition algorithms are based on featurebased methods. Some of them go through the process by detecting a set of geometrical features on the face such as distance between the eyes, eyebrows length, nose shape, and mouth width [1]. However, there is no distinction made between more or less prominent features, or analysis of how easy or hard it is to extract them. As an alternative, appearance-based face recognition algorithms are used as a tool to extract representation of an image by projecting it onto the subspace and then finding the closest point set [3]. O
Data Loading...
