Research Projects

Most of the work done in the lab has something to do with movement or temporal processing from modelling of the visual pathway to understanding dynamic changes in the face. 


Mutual GAZE TIming

Timing is extremely important for managing our social interactions. A glance or a stare can be uncomfortable. It turns out that people are most comfortable with mutual eye contact of around 3.3 seconds.

Binetti N, Harrison C, Coutrot A, Johnston A, Mareschal I. (2016) Pupil dilation as an index of preferred mutual gaze duration. R. Soc. open sci. 3: 160086. (PDF)



We know more about how we encode local information such as motion and colour but little about how we encode metric information like distance. The distance between two dots is not a given - the visual system has to work this out. In this study we show that apparent distance can be manipulated by adapting to a random dot texture.

Hisakata, R., Nishida, S. and Johnston, A., 2016. An Adaptable Metric Shapes Perceptual Space. Current Biology : CB. 26(14), 1911-5. (PDF)

Motion Modelling

Since 1992 we have been developing a model of human motion processing based on the calculation of spatio-temporal derivatives. The model takes image motion (top left) computes the local velocity represented in terms of direction (bottom right) coded with respect to the colour wheel) and speed (coded as brightness). 

Johnston, A, McOwan, P.W. and Benton, C.P. (1999) Robust velocity computation from a biologically motivated model of motion perception. Proc. R. Soc. Lond., B., 266, 509-518. (PDF)




Motion Grouping and the Harmonic Vector average

When an object moves at a constant velocity its motion is seen through the local analysis of neurones with small receptive fields. The result is that the the visual system can only encode the component of motion orthogonal to contours. We can study this psychophysically through an array of moving elements whose speed is related to local orientation in the same way e.g. in the example on the left vertically oriented contours don't move at all. We don't just average the motion as this would give a value that is half the true speed. However, for a unbiased set of elements the harmonic vector average provides the correct speed. The alternative is an approach called "intersection of constraints".

Johnston A and Scarfe P (2013) The role of the harmonic vector average in motion integration. Front. Comput. Neurosci. 7:146. doi: 10.3389/fncom.2013.00146 (PDF)


Time Perception

A common view is that our perception of time is governed by a single central clock. However we have shown that adaptation to flicker can change the apparent duration of a drifting stimulus in the adapted region of the visual field. The two test stimuli have identical durations. Keep you eye on the fixation spot when it moves. If you see a difference in apparent duration of the two test intervals this means the effect is retinotopic not spatiotopic.

Johnston, A., Arnold, D.H. and Nishida, S. (2006) Spatially localised distortions of event time. Current Biology, 16, 472-477. (PDF)


What are the elementary components of facial expression? We are studying interactions between moving facial features. In this study we found the movements of the mouth can influence the perceived speed of eye blinks. The eyelids appear to move slightly faster when the mouth stops moving. This shows that dynamic facial features are not coded independently.

Cook, R., Aichelburg, C., & Johnston, A. (2015). Illusory Feature Slowing: Evidence for Perceptual Models of Global Facial Change. Psychological Science,1–6. doi:10.1177/0956797614567340 (PDF)