PS1061: Sensation and Perception
Term 2,      THURSDAY 11am -1pm      (Boiler House)

Lecture 5:   Eye Movements and Perception

Lecturer: Szonya Durant, szonya.durant@rhul.ac.uk, (Room W 245)


Lecture topics


Useful links

http://viperlib.york.ac.uk/categories/121-eye-movements (requires you to register)

http://www.cudos.ac.uk/roger/oculo.html


Why study eye movements in Psychology?


Background - the eye and the retina

Images taken from https://www.stanford.edu/group/vista/cgi-bin/FOV/chapter-3-the-photoreceptor-mosaic

The most sensitive part of the retina is the fovea. Further away from the fovea neurons have larger receptive fields a poorer resolution.

eyeaxisTh e eye can move along 3 axes of movement.

Types of eye movements

Gaze shifting (orienting) mechanisms - Voluntary

  • Saccades - fast, ballistic, movements
  • Smooth pursuit - for tracking a moving object
  • Vergence - to adjust for the distance of the object to be viewed
Saccades - jerky movements for scanning the scene


pursuit
Smooth pursuit - following movements
vergence
Vergence - adjust to near/far objects

Stabilising mechanisms - Involuntary

  • Vestibulo-ocular reflex (VOR)- for correcting for head motion
  • Optokinetic reflex (OKR) - made in correction to the visual scene moving
spin
Spinning people round on a chair causes them to make eye movements to compensate- VOR
train
OKR can be seen when people look out of train windows - it manifests itself as nystagmus movie Some people experience nystagmus spontaneously, in this case it is considered a visual dysfunction http://www.nystagmus.co.uk/facts/the-basics

Further miniature eye movements

  • Microsaccades
  • Drifts and tremor
These small eye movements happen when we are fixing our eyes (fixating) on an object. You can read all about them here.

Measuring eye movements

  • There have been various ways developed, but most commonly now an eye-tracker is used.
  • This involves a small camera focused on the eye and software that tracks where the pupil moves.
  • By calibrating the pupil position against fixed points on the screen one can work out where the participant is looking.
 
trackers
Demo movie of eye tracking here

As well as the examples shown above, remote eye tracking (with no restriction on the head is now possible using two remote image sensors to capture images of the eyes and the reflection patterns. Using this kind of technology we can build a picture over time of where a participant was looking.

We can measure:
  • trajectory
  • frequency of fixation
  • dwell time
  • frequency of fixation
  • pre-define regions of interest
trackeye

How to represent and analyse eye movement data

Scanpaths and fixations

scanpict

Can simply trace the eye movements, or mark each fixation either with the time spent at
each point or with a circle that represents dwell time by its size.
scanpaths
Top image taken from Yarbus (1969), Chapter 7 here
Bottom image taken from eye tracker company Tobii website here: http://www.tobii.com

Heatmaps and regions of interest

heat
In this heatmap taken from Mosimann et al. (2004),
where the warmer colours represent more fixations - in this study they were looking at differences in attention in Alzheimer's sufferers.

roi
 In this study by Norbury et al. (2009) they chose regions of interest within movie clips of social
 interactions and compared the amount of fixations between different developmental groups within these regions e.g. eye, mouth, body.

Why do we need eye movements?

pigeon So we don't have to walk around like pigeons!!!
They don't have eye movements and bob their heads to compensate for their own motion and keep the image stable on their eye.
We use eye VOR and OKR to ensure stabilisation of the visual scene on the back of the eye (the retina).

To avoid fading of the visual scene, if a retinal image doesn't change neurons stop responding - eye movements act as a 'refresh'. This fading is called Troxler fading.

See a demo here: http://www.michaelbach.de/ot/col_lilacChaser/index.html for a great illusion using retinal after-images and Troxler fading.

ecc In order to focus the most important part of the scene on the fovea, the
part of the retina with highest resolution. The image on the left illustrates equivalent
size of letters according to our ability to see them if we are looking in the middle of the
image

Why don't we notice eye movements?

  • The brain needs to know which part of the motion on the retina is caused by eye movements and which part is caused by something moving in the scene.
  • This can be done by cancelling out the signal using an “efference copy” from eye muscles
  • Also, cues from the scene, such as whole image motion can be cancelled out.
  • Some illusions are caused by our inability to cancel out the motion of the eyes.
  • In these cases we can see our own eye movements (see image on right).

Image on right is taken from here: http://www.ritsumei.ac.jp/~akitaoka/index-e.html, where you

can see similar beautiful illusions. 

snakes

The importance of the active vision approach


Active vision emphasises the role of eye movements linked with attention in real-life tasks. Eye tracking studies such as Land et al. (1999) have measured how eye movements are made on to the next task, before the current one is completed.land

Eye movements during reading

  • Reading is complex cognitive process.
  • Many models exists to try and describe the ways in which we extract meaning from written words.
  • Eye movements are part of the evidence that can be used to test these models.
  • The brain needs to know which part of the motion on the retina is caused by eye movements and which part is caused by something moving in the scene.
  • They can also be used to manipulate the type of information present at fixation at any given time - called a gaze contingent paradigm
Saccades during test reading
reading
  • Rightward movements followed by return sweep
  • Large leftward movement to start of next line .
  • Fixations 200-250 ms, some regressions (10%) depending on text difficulty.
Aoccdrnig to a reresearchs at Cmabrigde Uinervtisy, it deosn't mttaer in waht oredr the ltteers in a wrod are, the olny iprmoatnt tihng is taht the frist and lsat ltteer be at the rghit pclae. The rset can be a toatl mses and you can sitll raed it wouthit porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe.
It has been shown that we do not fixate on each letter. Typically we move about 8 characters with each saccade. Many short words we do not even fixate on at all. The example above illustrates that reading does not rely on fixating each letter in order as we are still bale to make sense of the paragraph,
Gaze contingent paradigm
This involves displaying the text according to where the eye is looking e.g.:
Normal … move smoothly across the text…
Shown . …xxxxxxmoothly axxxxxxxxxx…..
Eye position ............*..............................
This can be used to see how many letters around fixation we use for reading.

The link between eye movements and attention

  • We are not able to process all visual information at once, we are only aware of a small part at any given time.
  • Usually where we are looking is where we are paying attention to (although not always).
  • The question is what makes us select something and how do we achieve this selection. How are eye movements and attention linked? This will be explored in Lecture 8.




Summary


Reading:

Specific References:


to download pdf-file of e-handout click here


Szonya Durant is not responsible for the content of external Internet sites

back to course outline
last update 11-02-2014
Szonya Durant