Signal Analysis: Chair Research Area Projects: - Biometric Person Authentication and Face Recognition - Content Based Queries for Image Databases Publications

Publikationer/
Publications

Program och
kurser/Programmes
and courses

Studentinformation/
Information for
students

Administration

Biometric Person Authentication and Face Recognition Project 

Objectives:

• To develop biometric analysis and decision tools for communications by using  Audio Visual data. 
• To develop methods for active "cameras'' with space variant visual reception fields (retina like organisation of the visual receptors) for the

purposes of Head and facial features tracking 
• Person authentication 

With the emergence of communication and computer science synergies, the interest for security issues is increasingly at the center of attention. An important element in this is automatic authentication of persons. Due to the lack of alternatives, much of the security in telecommunications applications, e.g. Internet, has been provided by encryption. In fact authentication and encryption are two different things which can not replace each other. For example, a safe transmission of a stolen or a given away Personal Identity Number (PIN) code does not guarantee that a remote service is used by the person it is intended for. For this reason, it is important to improve the technologies of authentication based on non-invasive biometric measurements such as those using video and audio. 

For communication it is important to 
accurately identify, and track faces, gestures, speech recognition. The latter can be improved via multi-modal ttechniques (e.g. a combination of lip movements and audio)

Applications:

Secure access and surveillance of 

buildings, rooms, offices 

terminals 

teleservices 

• Prof. Bigun has contributed to the European projects IT-VIRSBS and ACTS-M2VTS.
• Dr. Smeraldi has contributed to the European project of IT-VIRSBS.
• Tech. Lic. Nilsson has been contributing to the swedish national research consortium VISIT, via the project Image content retrieval for biometric applications 
• We promote common evaluation standards based on multimedia data e.g. M2VTS-Database, standardization, information diffusion on multi-modal person authentication.
• We are organizing the third International conference on Audio and Video Based Person Authentication, AVBPA.
• Take our face recognition test that attempts to quantify how humans perform in recognizing other humans. You can read the results for nearly 2000 persons in bigun01halmstad.pdf 

Publications
 

Back to Signal Analysis research area