Saturday, February 11, 2012

Profile Scanpath

This eyetracking scanpath of a bust of Nefertiti by Alfred Yarbus 50 years ago shows how one person's gaze surveyed this face in profile. 
It seems the ear is an important landmark, but the viewer didn't actually look at the cheek or chin very much. In an isolated object like this, it also appears that the eye is indeed following some contours, though in a jagged, jumpy sort of way.  Source of image


Jon Hrubesch said...

It seems to me that people are really just attracted to the areas of detail. I would assume with a cityscape the eyes will linger in the area of the buildings that have the most detail and spend little time in the sky and clouds. To me I would think as a viewer you're trying to absorb all the information in a picture to understand it. I'm sure if the picture were abstract scribbles with more concentrated areas a person would spend time in the concentrated areas even if they don't depict a physical identifiable object. At least that is what I'm taking away from these experiments.

Catherine said...

That's interesting - my own eyes were far more attracted to the planes of shadow on the cheek and neck, though the person observed here barely glanced at those areas.

Reminds me of a 2007 study comparing eye movements of laypersons and artists while looking at paintings. Non-artists tended to focus on areas of more detail (like people and faces), while artists were more likely to take in the entire picture, including negative space and the less-detailed areas.

I wonder if the person studied in this picture was trained as an artist? It doesn't seem to say in the source link.

Joseph Culp said...

i was wondering if they were an artist aswell, if that would account for the contour following

runninghead said...

I'd be interested to learn about the correlation between eye movement and the subjective apprehension of the gathered information too.
I mean, these are the points the physical eye dwells on, the cognitive 'eye' may well take that data, process it and present it to the concious mind in a slightly different manner.
Disparities between the two might contribute insights to our understanding of the whole process.

James Gurney said...

Runninghead, I think you've brought up a very important point, one that interests me, too. Put another way, these eye tracking experiments track foveal movement, but don't really measure how we perceive of larger shapes or contours through peripheral vision, or whether we're looking at the color or texture or thinking about something else entirely. I would suspect that one approach would be to correlate scanpath data with real-time brain scans. I've asked scientists in this arena how much we can deduce of cognitive experience from scanpath data, and they admit it's still somewhat limited.

phiq said...

@runninghead - Exactly what I was thinking; well said.

Erik Bongers said...

Hear, hear to runninghead.
One way that might help here is to let
artists view the image
The different pathway suggest different apprehension and cognitive processing.

Other factor is not only how often we focus on a detail but also how long we stay on it, and how slow we move to a next part, as is illustrated in
these more complex infographics. Staying longer on an item or moving slower to the next, might reveal more heavy cognitive processing.

Chris Beatrice said...

Call me skeptical, but I'm not sure that measuring precisely what the eye is aimed at (and for relatively how long) tells us all that much useful about picture-making or how we see.