EvoFIT: A Holistic, Evolutionary Facial Imaging Technique for Creating Composites

Charlie D. Frowd, Peter J.B. Hancock, Derek Carson

Research output: Contribution to journalArticleResearchpeer-review

92 Citations (Scopus)

Abstract

EvoFIT, a computerized facial composite system is being developed as an alternative to current systems. EvoFIT faces are initially presented to a witness with random characteristics, but through a process of selection and breeding, a composite is “evolved.” Comparing composites constructed with E-FIT, a current system, a naming rate of 10% was found for EvoFIT and 17% for E-FIT. Analysis revealed that target age was limiting factor for EvoFIT and a second study with age-appropriate targets visible during composite construction produced a naming rate similar to E-FIT. Two more-realistic studies were conducted that involved young target faces and two current systems (E-FIT and PROfit). Composites from both of these experiments were poorly named but a significant benefit emerged for EvoFIT.

Original languageEnglish
Pages (from-to)19-39
Number of pages21
JournalACM Transactions on Applied Perception
Volume1
Issue number1
DOIs
Publication statusPublished - 1 Jan 2004
Externally publishedYes

Fingerprint

Breeding
Composite
Imaging
Imaging techniques
Composite materials
Target
Face
Large scale systems
Profitability
Profit
Limiting
Alternatives
Experiments
Experiment

Cite this

@article{baab7ae5ac0244a5a4ed341cdc65efaa,
title = "EvoFIT: A Holistic, Evolutionary Facial Imaging Technique for Creating Composites",
abstract = "EvoFIT, a computerized facial composite system is being developed as an alternative to current systems. EvoFIT faces are initially presented to a witness with random characteristics, but through a process of selection and breeding, a composite is “evolved.” Comparing composites constructed with E-FIT, a current system, a naming rate of 10{\%} was found for EvoFIT and 17{\%} for E-FIT. Analysis revealed that target age was limiting factor for EvoFIT and a second study with age-appropriate targets visible during composite construction produced a naming rate similar to E-FIT. Two more-realistic studies were conducted that involved young target faces and two current systems (E-FIT and PROfit). Composites from both of these experiments were poorly named but a significant benefit emerged for EvoFIT.",
author = "Frowd, {Charlie D.} and Hancock, {Peter J.B.} and Derek Carson",
year = "2004",
month = "1",
day = "1",
doi = "10.1145/1008722.1008725",
language = "English",
volume = "1",
pages = "19--39",
journal = "ACM Transactions on Applied Perception",
issn = "1544-3558",
publisher = "Association for Computing Machinery (ACM)",
number = "1",

}

EvoFIT : A Holistic, Evolutionary Facial Imaging Technique for Creating Composites. / Frowd, Charlie D.; Hancock, Peter J.B.; Carson, Derek.

In: ACM Transactions on Applied Perception, Vol. 1, No. 1, 01.01.2004, p. 19-39.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - EvoFIT

T2 - A Holistic, Evolutionary Facial Imaging Technique for Creating Composites

AU - Frowd, Charlie D.

AU - Hancock, Peter J.B.

AU - Carson, Derek

PY - 2004/1/1

Y1 - 2004/1/1

N2 - EvoFIT, a computerized facial composite system is being developed as an alternative to current systems. EvoFIT faces are initially presented to a witness with random characteristics, but through a process of selection and breeding, a composite is “evolved.” Comparing composites constructed with E-FIT, a current system, a naming rate of 10% was found for EvoFIT and 17% for E-FIT. Analysis revealed that target age was limiting factor for EvoFIT and a second study with age-appropriate targets visible during composite construction produced a naming rate similar to E-FIT. Two more-realistic studies were conducted that involved young target faces and two current systems (E-FIT and PROfit). Composites from both of these experiments were poorly named but a significant benefit emerged for EvoFIT.

AB - EvoFIT, a computerized facial composite system is being developed as an alternative to current systems. EvoFIT faces are initially presented to a witness with random characteristics, but through a process of selection and breeding, a composite is “evolved.” Comparing composites constructed with E-FIT, a current system, a naming rate of 10% was found for EvoFIT and 17% for E-FIT. Analysis revealed that target age was limiting factor for EvoFIT and a second study with age-appropriate targets visible during composite construction produced a naming rate similar to E-FIT. Two more-realistic studies were conducted that involved young target faces and two current systems (E-FIT and PROfit). Composites from both of these experiments were poorly named but a significant benefit emerged for EvoFIT.

UR - http://www.scopus.com/inward/record.url?scp=85024289873&partnerID=8YFLogxK

U2 - 10.1145/1008722.1008725

DO - 10.1145/1008722.1008725

M3 - Article

VL - 1

SP - 19

EP - 39

JO - ACM Transactions on Applied Perception

JF - ACM Transactions on Applied Perception

SN - 1544-3558

IS - 1

ER -