Biometrics

Biometrics is the use of body or behavioural characteristic in order to identify a person or to verify his or her claimed identity.

At present, biometrics is receiving world-wide attention as a potential means to secure access to content and locations and to authenticate (internet) transactions. As it uses body (or occasionally behavioural) characteristics, it is the only known method capable of truly verifying or identifying a persons identity. Biometric recognition also has the potential of combining security with user convenience.

Five Camera experimental setup with controlled illuminationAlthough a number of recognition systems based on various types of biometrics such as fingerprint, human face, hand shape, iris pattern and voice are commercially available, biometric recognition is not yet generally accepted as a reliable building block in security systems. The reasons are:

  1. Lack of robustness: The performance of most of these systems degrades when used in real-life conditions.
  2. Sensor or condition variability: The performance of biometric technology can significantly degrade when a mismatch exists between the training condition and testing environment, either due to sensor differences or due to incompatibilities between training set and real-life conditions.
  3. User-generated variability: A biometric characteristic is strongly related to the physical state of a person. Therefore, it may be difficult to capture, leading to an enrolment problem. Also, the characteristic may be altered during a person's life time.
  4. Integration with security systems: Biometrics may solve a recognition problem, but it also creates problems of a technical nature such as: the secure storage of biometric data, secure enrolment, auditing, detection of abuse and privacy problems.
  5. Integration in ambient systems Today's biometric solutions often run on some server platform. In the near future, we expect that more distributed and mobile systems will evade the personal, the private and the work environment. In these environments biometrics can play a role in securing the access to systems (e.g. PDAs) and services in a unobtrusive way. Prerequisite for this to happen is that biometric recognition be embedded in distributed, small, low-power systems, even at the expense of a performance loss.
  6. User acceptance: This aspect is often neglected despite its primary importance to guarantee the success of a biometric application. For example, fingerprint recognition is sometimes perceived as linked with forensic investigation and may therefore be directly rejected by some users. Potential users may be reluctant to avail their biometric characteristic, fearing possible abuse.

Research projects 

  • 3D face reconstruction project: Person Verification 3D (PV3D).
  • The improvement of robustness of fingerprint recognition.
  • The IOP-GenCom project BASIS on biometric in the home environment. This involves research on transparent biometrics, not requiring specific user actions, and anonymous biometrics, aiming at avoiding the storage of biometric data in the clear.
  • The STW project Secure Grip. The main research question addressed in this project is whether an image of the pressure pattern exerted while holding an object can be used to reliably authenticate or identify a person. The work on grip pattern biometrics intended, among other things, to make police guns safer.
  • The Sentinels project ProBite. Although biometric identification and verification can be applied to enhance security, storing biometric data, known as templates, in a database introduces new security and privacy risks, which increase if the database is part of a network. The following threats can be identified: (a) Impersonation: an attacker steals templates from a database and constructs artificial biometrics that pass authentication. (b) Irrevocability: once compromised, biometrics cannot be updated, reissued or destroyed. (c) Privacy: exposure of sensitive personal information. A solution is to apply template-protection techniques, which make it impossible to recover the biometric data from the templates and enables revocation. The project's goals are to solve the problems of combining biometric identification and template protection and to validate the solutions in a home-network demonstrator. This demonstrator will be developed at Philips Research. It will connect devices such as DVD players, TV sets, etc. Fingerprint recognition will be used to identify the user and to control the access to content and devices. Template protection will be used to protect biometric data. The project is subdivided into three work packages. The first is dedicated to the development of template-protection methods. The second addresses the adaptations to fingerprint recognition required to use it in conjunction with template protection. In the third work package the home-network demonstrator will be developed.
  • The Freeband-BSIK project PNP2008. In the context of this project on personal networks, research is being done on transparent biometric verification for mobile devices.
  • The FP6 project3DFace. This is a European FP6 project with many European partners, like Fraunhofer, Sagem, University of Kent etc. It's aim is to develop 3D face recognition methods and 3D+2D combined methods for border control at airports etc. One of the biggest challenges in the project is that low error rates must be realised under realistic operational circumstances. Not only face regocgition itself is addressed, but also the development of a 3D face acquisition system and the integration into a complete system. SAS participates in work packages for 3D face recognition algorithms, development of a multiple view 3D scanner and fusion of 3D and 2D results.
  • The FP7 project TURBINE studies identity protection by means of biometric template protection for fingerprints.
  • The Marie Curie ITN Bayesian Biometrics for Forensics (BBfor2) provides a training infrastructure that will educate Early Stage Researchers in the core biometric technologies ofspeaker, face and ngerprint recognition, as well as the forensic aspects of these technologies.
  • The Itea2 project GUARANTEE provides a technical solution for personal safety in the home environment. GUARANTEE introduces local and network-supported decision making for safety applications on the basis of sensor input and with immediate response and feedback to the people concerned.
  • MMA (Multi-modal anthropometrics) This project studies person identification based on anthropometric measurements from surveillance cameras for forensic applications.
  • The NWO project FFR (Forensic Face Recognition) combines the elds of forensic face comparison and biometric face recogni-tion in order to develop a (partially) automated system for forensic facial comparison that quanti es the evidential value as a likelihood ratio thus supporting the court to make an objective decision.

People Members of Biometric Group present at ICB 2009 (left-right: Berk, Chun, Raymond, Anne, Bas)

Relevant calls for conference papers

Name

 Date

 Deadline

 Link

 Refereed/

Nonrefereed

IEEE WIFS 2012

Dec 02-05, 2012

Jul 01, 2012

http://www.wifs12.org

R

ICPRAM 2013

Feb 15_18, 2013

Jul 25, 2012

http://www.icpram.org/

R

VISAPP 2013

Feb 21-24, 2013

Jul 31, 2012

http://www.visapp.visigrapp.org

R

IEEE FGR 2013

Apr 22-26, 2013

Sep 15, 2012

http://fg2013.cse.sc.edu/

R
SPIE DS108
29 April - 3 May 2013 Oct. 22, 2012 http://spie.org/ds108
?
IWBF 2013
Apr 04-05, 2013
Nov. 9, 2012 http://www.img.lx.it.pt/iwbf2013/
?
CVPR 2013 June 23-28, 2013 November 15th, 2012 http://www.pamitc.org/cvpr13/home.htmlhttp://www.pamitc.org/cvpr13/home.html R
ICCV 2013 December 1-8, 2013 April 8, 2013 http://www.iccv2013.org/v R



    Last modified: 2013-02-28 (11:59) by Luuk Spreeuwers