Giving crime the 'evo': catching criminals using EvoFIT facial composites Charlie D. Frowd* 1 , Peter J.B. Hancock 2 , Vicki Bruce 3 , Alex H. McIntyre 2 , Melanie Pitchford 4 , Rebecca Atkins 1 , Andrew Webster 5 , John Pollard 6 , Beverley Hunt 7 , Emma Price 7 , Sandra Morgan 7 , Adrian Stoika 8 , Romeo Dughila 9 , Sergiu Maftei 9 and Gabriel Sendrea 9 1 Department of Psychology, University of Central Lancashire, Preston, UK 2 Department of Psychology, University of Stirling, Stirling, UK 3 School of Psychology, Newcastle University, Newcastle, UK 4 Department of Psychology, Lancaster University, Lancaster, UK 5 Force Major Investigation Team, Lancashire constabulary, UK 6 Family Support Unit, Lancashire constabulary, UK 7 Forensic Services Unit, Derbyshire Constabulary, UK 8 NASA Jet Propulsion Laboratory, NASA, USA 9 Forensic Department of Iasi County Police Inspectorate, Romania *Corresponding author: Dr Charlie Frowd, Department of Psychology, University of Central Lancashire, Preston PR1 2HE. Email: cfrowd@uclan.ac.uk. Phone: (01772) 893439. Abstract Facial composites are traditionally made by witnesses and victims describing and selecting parts of criminals’ faces, but this method is hard to do and has been shown to be generally ineffective. We have been working on an alternative system called EvoFIT for the past 12 years. Much of the development for this system has been carried out in the laboratory but, more recently, police forces have been formally evaluating it in criminal investigations. The current paper describes three of these police audits. It was found that EvoFIT composites made from real eyewitnesses led to an overall arrest in 23.4% of cases, which is similar to a 24.5% correct naming level found by laboratory research. System performance therefore appears to be effective both using simulated testing procedures and in the hands of the intended user. 0. Warning about content This technical paper contains descriptions of criminal offences. Some of these offences were against young people and are very serious. While individual case studies are not described in detail here, some readers may still find them disturbing; these readers may prefer to skip over the case studies: they are clearly labelled. 1. Introduction Witnesses and victims may be asked to construct a picture of a person they have seen commit a crime. This picture is known as a facial composite and can help the police to detect the perpetrator. Various techniques are available to produce these images. Traditionally, eyewitnesses describe the face in detail and build the composite by selecting individual facial features: hair, eyes, nose, mouth, etc. However, we do not perceive faces in such a piecemeal fashion, but instead process it more as a complete entity [1]. For this reason, face construction using the ‘feature’ approach often results in poor quality images, especially when carried out several days after the face had been originally seen [2], the normal situation for eyewitnesses. More recent so-called ‘holistic’ approaches [2-4] have attempted to overcome this problem using interfaces that require focus on the face as a whole. In essence, users repeatedly select complete faces from screens of alternatives, with cycles of ‘breeding’ in between, to allow a composite to be ‘evolved’. The approach has the potential to allow construction when a witness has seen the face, but cannot describe it in detail. It is in this situation that the traditional methods cannot be used effectively, since they require a good description for locating subsets of features (for presentation to the witness). In contrast, the ‘holistic’ systems only require fairly general information: age, gender, race and hair. The focus on the current paper is on one of these methods, EvoFIT. This system has been the focus of considerable research and development in the laboratory, and some of this work has been presented at previous BLISS conferences [5,6]. For three years now, EvoFIT has also been the subject of formal police field trials. These have explored the effectiveness of the system when 2010 International Conference on Emerging Security Technologies 978-0-7695-4175-4/10 $26.00 © 2010 IEEE DOI 10.1109/EST.2010.38 36