Why EvoFIT was developed

The construction of a composite by selecting individual facial features is common practice in criminal investigations. Photofit, identikit, efit and other ‘feature’ systems are old-fashioned and results are usually ineffective when used to identify suspects. This is why EvoFIT was developed.

Current practices may build composites with a single face, but operating on only one face means that a witness must continuously describe changes necessary to improve the likeness. This “recall” process is a naturally hard task for anyone to do. Recognising a face, on the other hand, is fairly easy and accurate. EvoFIT therefore attempts to tap into witnesses’ recognition rather than recall (research publications).

How does EvoFIT work?

Traditional composite systems used by police forces require witnesses and victims to describe an offender’s face and select individual facial features; tough, when under pressure.

EvoFIT uses a different approach. Faces in EvoFIT are not separated into component parts but modelled in their entirety (as a whole face). A facial composite is created by first displaying a number of faces containing random features such as eyes, noses and mouths. A witness selects a few of these faces that are most similar to a criminal. The selected faces are then mixed or ‘bred’ together to produce another set for selection. Repeating this process a few times allows a composite to be ‘evolved’.

EvoFIT begins by creating a set of faces with random facial shapes and facial textures. A witness would normally select six of these shapes and textures that most resemble the suspect. These selections become the “parents” of the next population; to produce another generation, the components of the selected faces are mixed together. The “offspring” faces are selected and combined together as before. The selections enable the set of faces to become more like the suspect. Evolution is completed when an acceptable likeness emerges.

This system is unique in helping a witness to focus on the central part of the face, the region that is important for recognition by another person later (a police officer or member of the public). The system also contains holistic tools to improve the match — 15 scales that change the overall appearance of the face such as perceived age, health, extroversion and masculinity. Other software tools are used adjust facial features for their shade, shape and position on the face. More recent work indicates the importance of the region around the eyes, a focus that is now part of the witness interview and face-construction procedure, as well as optimising the number of screens a witness sees to evolve the face.

The system can also allow a witness to construct the face using a self-administered procedure, an approach that aims to reduce police resources. This version is intended for use with volume (less-serious) crime.

Once the best likeness has been achieved, the image is saved as the EvoFIT composite. This image can then be circulated within the police, or more widely in the media, for identification. There are also various techniques that can be done to enhance the image for recognition, such as viewing the face as a dynamic caricature, as a stretched image, or in a number of “disguises”.

… from a mugshot album or an identity parade

A natural alternative is to present groups of faces and allow a witness to select a few based on their similarity to a suspect. Selecting in this way is rather like picking a criminal from a mugshot album or an identity parade. The task can be carried out without having to describe a face. What is required then is a method of combining these similar looking faces to provide an identifiable likeness. In spite of considerable obstacles, the system designers are delighted to have been able to achieve this aim. The research underpinning EvoFIT has been compiled by Professor Frowd and is available below as well as on Charlie’s website.

How was EvoFIT developed?

EvoFIT was developed in collaboration with Professors Peter Hancock at the University of Stirling (Department of Psychology), Vicki Bruce at Newcastle University (School of Psychology), and Charlie Frowd at the University of Central Lancashire (School of Psychology and Humanities), with funding from the Engineering and Physical Research Council (EPSRC) and the Department of Trade and Industry (DTI). Funding has also been gratefully received from the Higher Education Funding Council for England (HEFCE) and the North-West Development Agency (NWDA) that supported the Knowledge Exchange programme, Crime Solutions, at the University of Central Lancashire.

Contact Us

How can we be of assistance?