Biometrics-based security techniques - identification and verification using human characteristics and behavioural traits such as fingerprint and iris recognition - are playing an increasingly important role in securing information systems and physical assets, both in government and industrial contexts, especially in the aftermath of recent terrorist attacks, identity theft and internet-based fraud. There will be a long-term need for expert professionals trained to a high level in order to meet the rising market demand.
Research in biometrics has been affected by the lack of assessment tools and in particular the limited availability of public databases. Indeed accuracy evaluation of biometric algorithms is highly related to measuring statistics of errors on databases, which should be as large as possible. But, acquiring databases is very costly and constrained by data privacy rules. Moreover evaluation on proprietary databases does not allow any comparison between different algorithms. There is therefore a need of ressources for improving research innovation and education in the biometric domain.
The major role of "Association BioSecure” (BioSecure Foundation) is to maintain and distribute the major outcomes of the BioSecure FP6 Network of Excellence that lasted from June 2004 to September 2007.
It will provide to the biometric R&D community, resources such as:
Access to these resources is restricted to the members of the Association (see how to join).
-- LAST NEWS --
- New databases released: "Audio Video and still Face" and "multimodal-DS1" are available. More information here.
Reminder: All the databases can be ordered here.
- The 4th General Assembly took place on December 14th, 2012 in Nano-INNOV, Palaiseau, France in conjunction with the COST IC 1106 MC meeting.
- Results of the ESRA 2011 contest were presented in IJCB 2011.
- The new version of the iris reference system OSIRIS is available. In the lastest version (V4.1), the segmentation is based on the Viterbi algorithm, which is applied on the gradient map of images processed by Anisotropic Smoothing. The Viterbi algorithm is exploited at two resolutions: at a high resolution, it allows finding precise contours, while at a low resolution, the coarse contours are recovered. To improve the accuracy of the normalization step, the coarse contours are directly used in the Daugman's rubber-sheet to obtain the iris and pupil borders.
- The book about BioSecure Evaluation Framework has been available since April 2009: Guide to Biometric Reference Systems and Performance Evaluation by D. Petrovska-Delacretaz, G. Chollet and B. Dorizzi.
-- LAST NEWS --