Research Output

Catching More Offenders with EvoFIT Facial Composites: Lab Research and Police Field Trials

  Often, the only evidence of an offender’s identity comes from the memory of an eyewitness. For over 12 years, we have been developing software called EvoFIT to help eyewitnesses recover their memories of offenders’ faces, to assist police investigations. EvoFIT requires eyewitnesses to repeatedly select from arrays of faces, with ‘breeding’, to ‘evolve’ a face. Recently, police forces have been formally evaluating EvoFIT in criminal cases. The current paper describes four such police audits. It is reported that EvoFIT composites directly led to an arrest in 25.4% of cases overall; the arrest rate was 38.5% for forces that used a newer, less detailed face-recall interview. These results are similar to those found in the laboratory using simulated procedures. Here, we also evaluate the impact of interviewing techniques and outline further work that has improved system performance.

  • Type:


  • Date:

    01 April 2011

  • Publication Status:


  • Library of Congress:

    K1 Law (General)

  • Dewey Decimal Classification:

    364 Criminology


Frowd, C. D., Hancock, P. J. B., Bruce, V., Skelton, F. C., Atherton, C., Nelson, L., …Sendrea, G. (2011). Catching More Offenders with EvoFIT Facial Composites: Lab Research and Police Field Trials. Global Journal of Human-Social Science, 11(3),



Facial composite, witness, victim, EvoFIT, recognition, memory, interface, crime,

Monthly Views:

Available Documents

  • pdf

    Catching More Offenders With EvoFIT Facial Composi

    Number of Downloads in the past year: 1

    © 2011. Charlie D. Frowd, Peter J.B. Hancock, Vicki Bruce, Faye C. Skelton, Chris Atherton, Laura Nelson, Alex H. McIntyre, Melanie Pitchford, Rebecca Atkins,Andrew Webster, John Pollard, Beverley Hunt, Emma Price, Sandra Morgan,Roz Greening, Adrian Stoika, Romeo Dughila, Sergiu Maftei and Gabriel Sendrea. This is a research/review paper, distributed under the terms of the Creative Commons Attribution-Noncommercial 3.0 Unported License,
    permitting all non-commercial use, distribution, and reproduction inany medium, provided the original work is properly cited.

  • Downloadable citations