A face in the crowd: why facial recognition doesn't work like in the movies
Facial recognition systems are still not 100% accurate.
In recent years, my direction within psychology has led to the emergence of two related paths of research:
1. Face perception and social cognition - Utilising an evolutionary approach, I have been focusing on the signalling of personality and health information from the face, both in humans and chimpanzees, and have proposed the idea of a shared system across species. This investigation into social signals has also included own- and other-race faces, as well as information signalled through gait (using motion capture techniques).
2. Facial recognition and within-person variability - I have been using computational modelling in order to investigate the nature of within-person variability. I am trying to understand how we are able to recognise a familiar person from multiple (unstandarised) photographs, despite how varied these images often are. Through the use of principal components analysis and other techniques, I hope to model the variability of an individual and explore how idiosyncratic this variation might be.