Bilateral Australia: Solving the passport problem: Re-designing photo-ID to improve recognition

Lead Research Organisation: University of Glasgow
Department Name: School of Medicine

Abstract

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.

Publications

10 25 50
publication icon
Burton AM (2010) The Glasgow Face Matching Test. in Behavior research methods

publication icon
David White (Author) (2013) The crowd watches the crowd: Improving face matching performance by aggregating facial similarity ratings.

publication icon
Edmond G (2010) Atkins v The Emperor : The 'Cautious' Use of Unreliable 'Expert' Opinion in The International Journal of Evidence & Proof

publication icon
Jenkins R (2011) Stable face representations. in Philosophical transactions of the Royal Society of London. Series B, Biological sciences

publication icon
Jenkins R (2011) Variability in photos of the same face. in Cognition

publication icon
Megreya AM (2011) The other-race effect does not rely on memory: Evidence from a matching task. in Quarterly journal of experimental psychology (2006)

publication icon
White D (2014) Redesigning photo-ID to improve unfamiliar face matching performance. in Journal of experimental psychology. Applied

publication icon
White D (2013) Crowd Effects in Unfamiliar Face Matching in Applied Cognitive Psychology

publication icon
White D (2014) Feedback training for facial image comparison. in Psychonomic bulletin & review

publication icon
White D (2014) Passport officers' errors in face matching. in PloS one

 
Description The project has had three major scientific impacts.
Performance levels in face matching. Prior to this project, the study of face matching had been conducted mostly in the laboratory. Here we were able to compare face matching in the lab, with matching in real-world situations where people present their photos for verification. Similarly-poor levels of performance were observed in both settings, confirming that this is a significant problem for security.
Can we improve photographic ID? An explicit aim of the project was to establish whether matching performance could be improved by substituting average faces or cropped faces for standard ID photos. We observed a small but reliable advantage for both types of modified photo in the lab. However, this advantage was not observed in the real- world settings. This is an important divergence, and requires further investigation.
Can we improve the performance of viewers? In this project, we established, for the first time, that unfamiliar face matching can be significantly improved through training. We used a simple item-by-item feedback regime in a face matching task, and showed very significant improvement by viewers. This effect seems to have its source in viewers' over-estimation of their own performance - often assuming they have made a correct match when they have not. The feedback causes self-reflection, which in turn improves performance.
Exploitation Route Our research has had impact in two major ways.
Professionals using face matching. A great many professional groups rely, to some extent, on face matching to support security. The work here demonstrates the unreliability of unfamiliar face matching, and this is a message which is beginning to be understood in professional circles. However, it is important to observe that there are cultural issues involved in any professional practice, and we cannot claim to have changed professional behaviour overnight. Nevertheless, the importance of our findings is beginning to be understood by these groups.
Face matching as evidence. Psychologists have a long history of applying their findings to legal settings, and we had the opportunity during this project to collaborate with an international group of lawyers based in Australia (which has similar judicial procedures to the UK for identification evidence). This resulted in the article listed below, published in a legal journal, and reaching exactly the important target audience for this type of work. Our finding, essentially, is that human performance in identification is much less reliable than people have imagined - even when there is no reliance on memory.
Sectors Security and Diplomacy
 
Description The following article, published in a legal journal, represents a collaboration between psychologists and lawyers, arguing that evidence based on face matching techniques is highly unreliable. Edmond, G., Kemp, R., Porter, G., Hamer, D., Burton, M., Biber, K. & San Roque, M. (2010). Atkins v The Emperor: The cautious use of unreliable expert opinion. The International Journal of Evidence and Proof, 14, 146-166. As well as this published output, we have also disseminated findings from the project to the following organisations through meetings arranged for this purpose: Passport Offices at Canberra (July 2009) and Sydney (Sept 2009. Nov 2009) International Civil Aviation Organisation, Sydney (Sept & Oct 2009) Department of Foreign Affairs and Trade, Canberra (January 2010) 'Five Nations' Biometrics Working Group, Melbourne (January 2010) Defence Science and Technology Organisation, Adelaide (February 2010)
Sector Security and Diplomacy
Impact Types Societal