Open Source Platform for Biometric Authentification
The biometric user authentication yields to verify the identity of a person based on some physical or behavioural characteristics, such as fingerprints, DNA, iris, face or voice. This topic has been the object of a significant increase in interest during the last decade, in the commercial field (secured access to sensible information, secured transfer of personal information, identity verification for online transaction, entrance door control) or for identity verification enforcement (border crossing, biometric passport). Whereas the classical electronic identity verification techniques use a knowledge like a password to verify an identity or a possession like a key or a pin code, biometric authentication approaches associate the identity and the person him/herself. It enforces the link between the physical person and the claimed identity.
The MISTRAL project proposes an open source framework for biometric authentication. The first objective of MISTRAL is to facilitate the access to the biometric technologies for both academic researchers and teachers and for engineers from companies, providing them with a software toolkit, complete, powerful, with high flexibility, easy to understand and able users to deal with the majority of biometric applications on a large set of hardware environments. The second objective of MISTRAL is to promote biometric researches and applications by organizing communications and exchanges within a large set of MISTRAL software users, composed by actors coming both from academic and industrial worlds.
From a scientific point of view, the main originality of MISTRAL is to propose a unique recognition engine for different modalities (like voice and face) and different applicative contexts, like linux/windows computers, distributed environments, telecommunication environments, embedded systems (PDA and cell phones) and smart rooms. This characteristic offers several advantages, like focusing the research work on the specificities of one biometric modality, helping the fusion of modalities (multimodal systems would play a key role in biometrics in the next few years), reducing the development effort for a new application… One goal associated with MISTRAL is to be useful for fast and efficient designing of commercial demonstrators and for adding easily a new modality to a system, for research or applicative purposes.
People involved in SAMOVA team
- Jérôme Farinas (scientific coordinator)
- Eduardo Soto-Sanchèz
- Start time : 1st march 2007
- End time : 1st march 2009