Automatic decoding of facial movements reveals deceptive pain expressions
Facial expressions are an important source of information for social interaction. The subcortical extrapyramidal motor system is thought to be mainly responsible for spontaneous facial expressions and the cortical pyramidal system for voluntary experienced expressions of emotions.
The aim of this study was to test the abilities of both human observers and a computer vision system to discriminate between real or faked facial pain expressions. 170 human observers were shown clips of individuals experiencing both genuine pain while submerging their arm in ice water, or pretending to be in pain whilst submerging their arm in warm water. In the second experiment, 35 new participants were trained to asses weather this could improve their accuracy.
The results indicate that human observers only have a 60% accuracy rate to discriminate real expressions of pain from faked expressions, even after training. The computer vision system had an 85% accuracy. Facial expressions are only one element of the pain experience. However, these results could have clinical and practical implications in situations where the behavioural fingerprint of neural control systems may be relevant. > From: Bartlett et al., Curr Biol 24 (2014) 738-743. All rights reserved to Elsevier Ltd.
Visit the Pubmed summary for more information or your article access.