Using perceptual learning to understand and influence face recognition

Perceptual learning is a fundamental cognitive skill. It can be defined as an enhancement in the ability to distinguish between similar stimuli (that otherwise would be very hard to tell apart) as a consequence of experience with them, or with stimuli similar to the target stimuli. The proposed project focuses on the development of this phenomenon as a key factor to our ability to recognise faces. Further, by using a range of neuroscience techniques (EEG/ERP, tDCS, fMRI, TMS) in conjunction with the behavioural designs I have developed, I will investigate methods to improve this perceptual skill and define specific brain structures responsible for the control and development of this phenomenon. So, basically, I aim to find out how we improve at telling things apart (discrimination) and to discover ways of enhancing this ability.

I will start by developing the case for perceptual learning as a key contributor to one of the most robust cognitive phenomenon in face recognition i.e. the composite face effect. This refers to individuals' decreased ability to recognise the top half of one face presented in composite with the bottom half of another face when the composite is upright and aligned than when the two halves are offset laterally (misalignment). By using novel categories of prototype-defined chequerboards which participants will be pre-exposed to during the study procedure, I would expect to show a similar composite effect for familiar chequerboards to that usually found with faces. This will be our index of perceptual learning.

The project will then continue by using different neurostimulation techniques (tDCS and TMS) to selectively increase and decrease the composite effect for chequerboards and that for faces. Through a combination of tDCS and EEG techniques, this project will also reveal how specific brain responses usually found for faces, can also be found for familiar chequerboards and be altered by neurostimulation. Finally, a combination of tDCS and fMRI will help to localise more precisely the brain structures involved in perceptual learning.

This project will provide us with insights into the mechanisms that characterise the development and control of perceptual learning and will extend perceptual learning to face recognition by showing how similar effects to those usually found for faces can be found for categories of stimuli that participants had never seen before entering the lab. It will also provide evidence for the neurocognitive basis of perceptual learning. We will use these results to devise methods of training that can enhance our discrimination abilities in various domains. It would allow us to tailor training programmes (e.g. involving stimuli such as faces, fingerprints, cervical smears, mammograms) to provide maximum benefit in terms of performance by the trainee. As an example, this research would be suitable for enhancing the face recognition skills of frontline and intelligence staff, and one might be able to selectively increase the effectiveness of this training through the use of mobile neuroscience techniques based on the Starstim EEG/tDCS system. It would also give us insights into how to design artificial systems for the detection and recognition of such stimuli, which would be of interest to companies specialised in computer software or apps for face recognition.

This new approach for studying the mechanisms of perceptual learning using a combination of experimental psychology designs and neuroscience techniques will benefit several academic and non-academic organisations within and outside the UK.

It will create a database of advanced knowledge about perceptual learning that could serve several researchers as the basis for future research on human cognitive skills, neuroscience techniques, and the creation of computer software for discrimination and recognition of various stimuli (e.g. faces). This would, in turn, enable the PI and others to apply this body of knowledge to real-world problems such as face recognition both by humans (e.g. "super recognisers") and by machines. To this end, the PI and the Co-I already have initial links to Dr Josh Davis' lab at the University of Greenwich where they work with the Metropolitan Police Force on super recognisers, and links with Prof. Dominic Dwyer's lab at Cardiff where they are investigating the application of models of human perceptual learning to machine recognition of images in conjunction with a number of industrial partners.

Through the University of Exeter Impact Innovation & Business (IIB) service and ESRC Impact Accelerator Account, the PI will seek to establish new collaborations with public and private organisations so as to maximise the societal impact of the findings produced by the proposed research. This could benefit government agencies such as Defence (DSTL), Customs and Excise (screening for contraband) and some medical applications (e.g. radiography). One example, contributing to the work done by DSTL, would be training the military, in discriminating similar targets in combat situations. An application that could be developed as a consequence of this programme of research would be one that selectively increases the effectiveness of this training through the use of mobile neuroscience techniques based on the Starstim EEG/tDCS system. As another example, it could also benefit private organisations like IBM in developing new computer and phone apps to provide face and object recognition skills for medical testing (e.g. detecting early signs of neurodegenerative disorders affecting face or object recognition).

The PI will also attend the University of Exeter annual Research Focus Week event (2018-19-20) usually organised in May. Many of the sessions planned for this event, such as those on UK Government and Industry funding; EU funding, Engaged Research, Large Grants, Programme Grants and Centre Bids will allow the PI to discuss with the Exeter technical research staff and guest speakers further opportunities for him to use this research and so increase the impact of it within and outside the UK.

To engage with a wider public, the PI will present this research at the events organised by the University of Exeter that involve high-school students and their parents (National Science Week demonstrations and talks, School events). Also, he will present this research at research seminars within the Department of Psychology so that students from the different psychology domains within the department (social, clinical, cognitive and animal behaviour) can benefit from the research. The PI will also present this research at the Lecture and Research events organised by the University and open to the public.

Depositing the data and outputs from the proposed research on open access (through the Exeter ORE system) will ensure anybody from the public will be able to benefit from this research. The University Press Office will also help to extend my research findings to the public through the use of local, national and international newspapers, magazines, and online media.