Transparency of Geometric Distortions in Face Identity

Research has suggested that processing the spatial interrelationship of facial features is crucial to face perception (Diamond & Carey, 1986; Maurer, Le Grand, & Mondloch, 2002). However, recent research that has examined the effect of stretched and sheared images has suggested that early vi...

Full description

Bibliographic Details
Main Authors: A Sandford, A M Burton, D White
Format: Article
Language:English
Published: SAGE Publishing 2013-10-01
Series:i-Perception
Online Access:https://doi.org/10.1068/ig4
id doaj-68e9db4f3cf645b48fc58af589af1534
record_format Article
spelling doaj-68e9db4f3cf645b48fc58af589af15342020-11-25T02:54:29ZengSAGE Publishingi-Perception2041-66952013-10-01410.1068/ig410.1068_ig4Transparency of Geometric Distortions in Face IdentityA Sandford0A M Burton1D White2University of Aberdeen, UKUniversity of Aberdeen, UKUniversity of New South Wales, AustraliaResearch has suggested that processing the spatial interrelationship of facial features is crucial to face perception (Diamond & Carey, 1986; Maurer, Le Grand, & Mondloch, 2002). However, recent research that has examined the effect of stretched and sheared images has suggested that early visual processes do not discriminate between an undistorted image and a severe geometric distortion of the same image (Bindemann, Burton, Leuthold, & Schweinberger, 2008). In two experiments, we show that both familiar and unfamiliar face recognition may be robust to geometric distortions. Using a familiar/unfamiliar judgment task, Experiment 1 compared correct reaction times and accuracy for undistorted face images to vertically and horizontally stretched front-view images. In Experiment 2, we compared undistorted face images to stretched, sheared, and twisted face images seen in front- and three-quarter-viewpoints. Observers saw all familiar and unfamiliar faces randomly in all distorted conditions. In both experiments, we found robust accuracy in all conditions despite effects on correct reaction times and substantial within-image deformation in distorted conditions. The findings suggest that processing the spatial interrelationship of features may not be crucial for correctly recognising a familiar face and correctly rejecting an unfamiliar face. This raises questions about the role of configural information in face processing for identity.https://doi.org/10.1068/ig4
collection DOAJ
language English
format Article
sources DOAJ
author A Sandford
A M Burton
D White
spellingShingle A Sandford
A M Burton
D White
Transparency of Geometric Distortions in Face Identity
i-Perception
author_facet A Sandford
A M Burton
D White
author_sort A Sandford
title Transparency of Geometric Distortions in Face Identity
title_short Transparency of Geometric Distortions in Face Identity
title_full Transparency of Geometric Distortions in Face Identity
title_fullStr Transparency of Geometric Distortions in Face Identity
title_full_unstemmed Transparency of Geometric Distortions in Face Identity
title_sort transparency of geometric distortions in face identity
publisher SAGE Publishing
series i-Perception
issn 2041-6695
publishDate 2013-10-01
description Research has suggested that processing the spatial interrelationship of facial features is crucial to face perception (Diamond & Carey, 1986; Maurer, Le Grand, & Mondloch, 2002). However, recent research that has examined the effect of stretched and sheared images has suggested that early visual processes do not discriminate between an undistorted image and a severe geometric distortion of the same image (Bindemann, Burton, Leuthold, & Schweinberger, 2008). In two experiments, we show that both familiar and unfamiliar face recognition may be robust to geometric distortions. Using a familiar/unfamiliar judgment task, Experiment 1 compared correct reaction times and accuracy for undistorted face images to vertically and horizontally stretched front-view images. In Experiment 2, we compared undistorted face images to stretched, sheared, and twisted face images seen in front- and three-quarter-viewpoints. Observers saw all familiar and unfamiliar faces randomly in all distorted conditions. In both experiments, we found robust accuracy in all conditions despite effects on correct reaction times and substantial within-image deformation in distorted conditions. The findings suggest that processing the spatial interrelationship of features may not be crucial for correctly recognising a familiar face and correctly rejecting an unfamiliar face. This raises questions about the role of configural information in face processing for identity.
url https://doi.org/10.1068/ig4
work_keys_str_mv AT asandford transparencyofgeometricdistortionsinfaceidentity
AT amburton transparencyofgeometricdistortionsinfaceidentity
AT dwhite transparencyofgeometricdistortionsinfaceidentity
_version_ 1724720890116046848