DakinLab
UCL Institute of Ophthalmology
DakinLab
UCL Institute of Ophthalmology
Crowding. Fixing your eye on each number in turn, try and recognise the central object in the groups either side of where you’re looking. It’s harder on the left than on the right because of crowding which limits our ability to recognise objects in the visual periphery. The more similar an object is to it’s surroundings, the worse crowding is. Crowding is a problem for people who lose central vision and have to rely on their periphery
“Number-sense”. Patches circled in red and yellow contain the same number of dots but this is hard to tell. Similarly the spacing of dots in yellow and blue circled patches is the same but they appear more densely packed in the yellow region. It’s not until we add ~40% more (green-circle) that their spacing looks matched to the dots in the yellow-circle. This shows both our sense of spacing (density) and number are affected by the size of the region and likely are measured using the same visual information.
Face processing gets lots of information from horizontal information
(“barcodes”)
Lightness perception The dark and light regions in the face on the left are an illusion; masking out the contours (bottom-left) reveals that the regions are the same lightness. This effect is due to the “coarse scale” structure: when we scramble it (top right) the illusion disappears and when we scramble the fine structure (bottom right) the illusion persists. We have shown this arises from the visual brain expecting the statistical structure of images to be “natural”
Lightness perception II In the stripy image on the left the grey bars in the centre of the red rings are identical lightness but look very different! Pulling these patches “out of context” (right) shows they are the same. Again, guided by the statistics of the natural visual environment, we think is the brain is “factoring out” what looks like a shadow falling across the stripes.
Vision in schizophrenia Look at the region circled in red. When asked to pick a blue-circled patch that matches it’s appearance most people pick something around 20% even though it’s physically matched to the 40% patch(!) The high contrast surround makes the red-circled patch appear lower contrast. People with schizophrenia are much less vulnerable to this illusion which we think arises because of the disease leading to weaker inhibition (i.e. an inability of neurons to reduce one another’s activity)
40%
10%
30%
20%
Current areas of interest
Illustrative examples of previous and current work from my lab are below, but other things we are currently working on are:
•Metamorphosia in macular disease
•Early diagnosis of glaucoma
•Novel therapies for amblyopia
•Seeing motion
•Non-representational art and visual neuroscience