Observations during the 1970s of witnesses mis-remembering unfamiliar people from crime scenes has led to a lot of investigation into face recognition over the years. Today's rapidly increasing use of CCTV images makes the subject as topical as ever.
Previous research has shown that comparing images of unfamiliar faces to see whether they show the same person is highly prone to error, even when the pictures are high quality and the faces are shown at the same time to avoid relying on memory. By contrast, recognising familiar faces can be highly accurate, even when the image quality is extremely poor.
The aim of a new study led by Professor Vicki Bruce, now at Edinburgh University, was to investigate how faces become familiar to us. Her research is published today as part of the ESRC's Social Science Week. The study involved collaboration between researchers at the Universities of Stirling and Glasgow. "There has been good theoretical progress about the way we recognise familiar faces, " says Professor Bruce, "but there remains a gulf between our understanding of familiar and unfamiliar face perception and memory."
Researchers used a variety of experiments, involving volunteers viewing video sequences of people, watching episodes of unfamiliar soap operas (from Irish TV), and learning the visual appearance of familiar but previously unseen characters from radio's The Archers and voices from The Simpsons. They found that for unfamiliar faces, memory appears dominated by the 'external' features, but where the face is well-known it is 'internal' features such as the eyes, nose and mouth, that are more important.
The research team was able to trace shifts in the importance of internal features during the course of learning, and these shifts occurred rapidly - after just a few minutes of exposure.
Other experiments showed that people are better or more confident at learning faces which belong to well-known, though previously unseen, personal identities. In studies using faces claimed to belong to characters playing in The Archers and the Simpsons, learning new faces was found to be easier.
Said Professor Bruce: "The work we have done to date has been confined to the more visual end of learning, but the results have been striking. Our results lead us to think that mental images of faces come together relatively quickly - that it is minutes rather than hours of exposure that may be critical. They also show how visual learning can be supported by other knowledge of personal identities."
The studies found that the way in which faces were shown to volunteers -as static or moving images, or with or without accompanying voices - made no difference to how quickly they learned to recognise them. It appeared that the sum of the images was the most important factor, and computer simulation studies provided some clues about why this might be the case. Finally, research into identification using low quality CCTV images confirmed earlier findings of the ease with which this can be done for highly familiar faces, but also showed a gradual improvement as faces became better known.
For further information:
Contact Professor Vicki Bruce on 0131-650-4089, e-mail: Vicki.Bruce@ed.ac.uk
Professor Mike Burton on 0141-330-4060, e-mail: firstname.lastname@example.org
Or Lesley Lilley or Anna Hinds at the ESRC Press Office, on 01793 413119/413122
NOTES FOR EDITORS
1. Social Science Week 2003, from the 23rd to the 27th June, is about revitalising policy by bringing social scientists and their research together with policy-makers. Events in various locations will showcase a broad array of ESRC research. Topics will cover a wide spectrum, from the state of UK business to climate change and arms control. For a programme visit http://www.
2. The research report 'Getting To Know You - How We Learn New Faces' was funded by the Economic and Social Research Council (ESRC). Professor Bruce is now at the College of Humanities and Social Sciences, University of Edinburgh, EDINBURGH EH8 9JU.
3. The ESRC is the UK's largest funding agency for research and postgraduate training relating to social and economic issues. It has a track record of providing high-quality, relevant research to business, the public sector and Government. The ESRC invests more than £53 million every year in social science research. At any time, its range of funding schemes may be supporting 2,000 researchers within academic institutions and research policy institutes. It also funds postgraduate training within the social sciences thereby nurturing the researchers of tomorrow. The ESRC website address is http://www.
4. REGARD is the ESRC's database of research. It provides a key source of information on ESRC social science research awards and all associated publications and products. The website can be found at http://www.