International Biometric Society
British Region

Annual General Meeting and 208th Ordinary Meeting
``Biometry and Biometrics - the interaction between Statistics/Mathematics and Bioidentification''

This year the Autumn Meeting of the British Region of the International Biometric Society will revert to the recent format of a full-day meeting, incorporating the Annual General Meeting. This meeting will be held at 11am on Thursday 11th November in the Lecture Hall at the Royal Statistical Society, 12 Errol Street, London, EC1Y 8LX.

This meeting will start with the AGM which will be followed at approximately 11.20 by the Presidential Address of our retiring President, Prof. Mike Kenward ``What’s missing? Principles and procedures for handling missing data''. As is our usual practise we have two invited discussants for this address.

Following lunch (12.30 - 13.30) we have arranged a meeting devoted to the interaction between the two different uses of the term `biometrics', as given in the title above. This is a topic of current interest both within the UK and internationally, and at an international level, the IBS is currently establishing an ad-hoc committee to consider links and synergies between the IBS and `the other biometrics'. We have invited speakers covering this topic from a number of different angles, including statisticians and mathematicians, and also two speakers from the International Association for Biometrics (AfB), one of the professional bodies for `the other biometrics'. We are publicising this meeting both within the IBS and within the AfB, and hope that the audience will be a mixture of people from both organisations.

The full programme for the meeting is given below, together with short abstracts for the papers to be given during the meeting. The Agenda for the Annual General Meeting follows at the end of the programme.

Lunch will be available at the RSS for those attending both parts of the meeting. If you wish to have lunch it will be necessary to register in advance (registration form included with this newsletter) at a cost of £15 (same cost for members and non-members). Registration forms must be received by Thursday 4th November (address on registration form).

Information on how to get to the Royal Statistical Society can be obtained from their web-site http://www.rss.org.uk/about/direction.html.

Programme

11:00 – 11:20 Annual General Meeting (for IBS British Region Members only)

Agenda available separately

11:20 – 12:30 Presidential Address: What's missing? Principles and procedures in the handling of missing data.

Mike Kenward (London School of Hygiene and Tropical Medicine)

Until quite recently missing data was considered largely as a computational issue. However, in the past decade or so there has been an explosion in research on this subject and today about 20% of Biometrics associate editors name missing values as one of their topics of specialization. Areas of application range from clinical trials to surveys and epidemiology. This talk is aimed at those who have little knowledge of the subject, with the purpose of mapping out the main strands of development, providing, where possible, common links and themes. There are many different views of the subject, some strongly opposed. An attempt will be made to give an even-handed overview, with a stress on the distinction between principles and procedures in statistical practice.

Discussants: Geert Molenberghs (Limburgs University, Belgium)
Gillian Raab (NapierCollege)

12:30 – 13:30 Lunch (available at RSS if booked in advance – see registration form)

13:30 – 13:50 Biometry and Biometrics – the interaction between statistics/mathematics and bioidentification: An Introduction

Sheila Bird (MRC Biostatistics Unit, Cambridge)

13:50 – 14:25 Statistical Facial Identification

Nick Fieller (University of Sheffield)

There are many existing systems for automatic facial recognition which select the best available match to a questioned image of a human face to one or more images selected from a database of known people. These systems are successful and widely used in areas such as security surveillance. However, they do not attempt to provide any quantitative measure of quality of match but only give the best available match. In this respect they fall short of a facial identification which can give evidential information of use in a court of law. The work described here provides a statistically based method which can remedy these defects. It is based on landmark identification of facial features and routine techniques of shape analysis to provide measures of inter and intra variability of measured facial features, thus allowing a more statistical assessment of facial identification.

14:25 – 15:00 Statistics with a human face

Adrian Bowman & Mitchumn Bock (University of Glasgow)

In a longitudinal study of facial growth in babies and young children, stereophotogrammetry has been used to construct three-dimensional models of each face. This allows data in the form of landmarks, anatomical curves and surface meshes to be extracted for analysis. Techniques for display and exploration of the data will be described, with particular interest in comparing cleft lip repair cases with a sample of controls. Asymmetry is also of particular interest in this application and techniques for the measurement and analysis of this important feature will be discussed.

15:00 – 15:35 The interplay of perceptual and cognitive elements in fingerprint identification: When higher-level cognition can facilitate or hinder fingerprint matching.

1Dave Charlton, 2Ailsa Peron, and 2Itiel Dror (1Sussex Police Fingerprint Bureau; 2University of Southampton)

Fingerprint identification, like many other identification processes, is not totally dependent on the bottom-up perceptual elements. These elements provide the critical data for the examination of friction ridge skin detail. However, the processing, evaluation, and interpretation of this data will be highly dependent on a variety of higher-level cognitive elements. These higher-level elements can facilitate the processing of information, and can even enable a trained examiner to make a correct identification when perceptual information is degraded or even missing. In other cases, the higher-level elements can hinder correct identification. In this presentation we will present some of the perceptual and cognitive elements involved in fingerprint identification and how they work (or not) well together.

15:35 – 15:55 Tea

15:55 – 16:30 Recognising Persons by their Iris Patterns

John Daugman (University of Cambridge)

Iris recognition provides real-time, high confidence identification of persons by analysis and comparison of the complex patterns that are visible within the iris of an eye from some distance. Because the iris is a protected, internal, organ whose random texture is epigenetic and stable throughout life, it can serve as a living password or passport. Recognition decisions are made with confidence levels high enough to support exhaustive searches through national databases. The principle underlying the iris recognition algorithms is the failure of a test of statistical independence on phase sequences as encoded by multi-scale quadrature wavelets. This test of independence involving about 250 degrees-of-freedom is passed whenever different irises are compared, but it is failed when images of the same iris are compared. It is unnecessary for users to assert any identity which is then merely verified; rather, their identity is discovered by searching databases of enrolled IrisCodes at speeds of about 1 million persons/second/CPU. Data will be presented in this lecture from 9.1 million IrisCode comparisons and from independent test reports. These algorithms are used now for passenger identification at several airports and in related biometric applications.

16:30 – 17:05 Building a UK Centre of Excellence for Biometric Technologies

Clive Reedman (International Association for Biometrics)

The science base has made a pioneering contribution to biometrics technology over many years. UK leadership is recognised internationally in research into multi-modal biometrics, gait analysis and speaker recognition, interface design issues & evaluation of systems. There is a need to create a more effective collaborative network of universities to tackle more complex large-scale projects in the UK & internationally. A detailed study of biometrics technology has confirmed that it is a key strategic technology which the UK needs to develop. Biometrics is used to identify and authenticate individuals based on physiological or behavioural characteristics such as face, finger, hand, iris, voice signature or keystrokes dynamics. The rapidly growing market, the impact of 9/11 and the critical demands of governments require highly inventive new approaches to innovate and adapt the technology. To be successful, this will need enhanced collaboration between the science base, industry and Government. To facilitate this, the International Association for Biometrics, the University of Kent and Atmaana Consultants have been working hard for over a year to bring the centre into existence. Now, with the support of the DTI and Home Office, we are close to realising the objective.

17:05 – 17:15 Close