The Selective Tuning Model of Visual Attention

1
An Introduction toEye-Movement Research Marc
Pomplun Department of Computer
Science University of Massachusetts at
Boston E-mail marc_at_cs.umb.edu Homepage
http//www.cs.umb.edu/marc/
2
An Introduction toEye-Movement Research

• Overview
• Types of Eye Movement
• Why Eye-Movement Research?
• How to Measure Eye Movements
• Examples of Eye-Movement Studies and Paradigms

3
Eye Movements
4
Types of Eye Movement

• Fixations
• The eye is almost motionless, for example, while
  reading a single, short word.
• The information from the scene is almost entirely
  acquired during fixation.
• Duration varies from 100-1000 ms, typically
  between 200-600 ms.
• Typical fixation frequency is about 3 Hz.
• Fixations are interspersed with saccades.

5
Types of Eye Movement

• Saccades
• Quick jumps that connect fixations
• Duration is typically between 30 and 120 ms
• Very fast (up to 700 degrees/second)
• Saccades are ballistic, i.e., the target of a
  saccade cannot be changed during the movement.
• Vision is suppressed during saccades to allow
  stable perception of surroundings.
• Saccades are used to move the fovea to the next
  object/region of interest.

6
Types of Eye Movement

• Smooth Pursuit Eye Movements
• Smooth movement of the eyes for visually tracking
  a moving object
• Cannot be performed in static scenes
  (fixation/saccade behavior instead)

7
Types of Eye Movement

• Torsional Eye Movements
• Rotation of the eye around the viewing axis
• Stabilization of visual scene by compensating
  body rotation (up to about 15 degrees)

8
Types of Eye Movement

• Vergence Eye Movements
• Slow, smooth movements changing the vergence
  angle (the angle between the two viewing axes)
• Used for changing gaze from a near to a far
  object or vice versa
• Can take up to one second
• Execution is often interrupted if no thorough
  inspection of the object is required.

9
Types of Eye Movement

• Tremor
• Fast, low-amplitude (seconds of arc) eye-movement
  jitter
• Improves the perception of high spatial
  frequencies
• Prevents the fading of static images during
  fixations

10
Why Eye-Movement Research?

• Throughout my talks, the only types of eye
  movement that we will consider are fixations and
  saccades.
• Why?
• These eye movements indicate a persons gaze
  trajectory while performing a certain task.
• Moreover, they yield information about a persons
  visual attention.

11
Why Eye-Movement Research?

• About eye movements and visual attention
• Usually, saccades follow shifts of attention to
  provide high acuity at the attended position.
• It is possible to look at an object without
  paying attention to it (staring).
• It is possible to shift attention without eye
  movement (covert shifts of attention).
• It is impossible to perform a saccade while not
  shifting attention.
• During specific, natural tasks it is reasonable
  to assume that saccades follow shifts of
  attention.

12
Why Eye-Movement Research?

• The investigation of visual attention, in turn,
  is at the core of cognitive science.
• Studying visual attention yields insight into
  general attentional mechanisms.
• It can provide information on a persons
  stream of conscious and unconscious
  processing while solving a task.
• Attention is closely linked to the concept of
  consciousness.
• Attentional mechanisms could improve
  artificial vision systems.

13
How to Measure Eye Movements

• Mirror on Eyeball
• Used in first eye tracking experiments (Yarbus in
  1960s)
• Suction cup attaches mirror to eyeball
• Light beam is directed at mirror and reflected
  onto photo sensitive paper
• Good spatial resolution but no temporal
  information
• Unpleasant for the subject

14
How to Measure Eye Movements

• Electrooculogram (EOG)
• Skin electrodes around the eyes measure
  potential differences
• Wide range -- poor accuracy
• Better for relative than absolute eye movements
• Mainly used in neurological diagnosis

15
How to Measure Eye Movements

• Eye Coils
• Subject wears contact lens with wire coil
• Homogeneous magnetic field around subject allows
  measurement of gaze angle
• Very high temporal and spatial precision
• Special coil also allows measurement of torsional
  eye movements
• Very uncomfortable

16
How to Measure Eye Movements

• Limbus Tracker
• Photo diodes track the boundary between sclera
  and iris.
• High temporal resolution
• Poor spatial precision only for horizontal eye
  movements
• Inexpensive

17
How to Measure Eye Movements

• Purkinje Eye Tracker
• Laser is aimed at the eye.
• Laser light is reflected by cornea and lens
• Pattern of reflected light is received by an
  array of light-sensitive elements.
• Very precise
• Also measures pupil accomodation
• No head movements

18
How to Measure Eye Movements

• Video-Based Systems
• Infrared camera directed at eye
• Image processing hardware determines pupil
  position and size (and possibly corneal
  reflection)
• Good spatial precision (0.5 degrees) for
  head-mounted systems
• Good temporal resolution (up to 500 Hz) possible

19
How to Measure Eye Movements

• EyeLink II System
• Binocular
• Head-movement compensation (head camera looking
  at IR markers at monitor)
• Temporal resolution 500 Hz
• Spatial precision about 0.5 to 1 degree
• Gaze-position data available in real-time

20
How to Measure Eye Movements

• EyeLink II System Configuration

21
How to Measure Eye Movements

• Measuring vergence eye movements in anaglyphs
  with EyeLink II (poor subject!)

22
Eye-Movement Studies

• Eye movements while watching a girls face
  (early study by Yarbus, 1967)

23
Eye-Movement Studies

• Visual scan paths on instruments/dashboards
  studies for the improvement of human-computer
  interfaces

24
Eye-Movement Studies

• Gaze trajectory measurement for the optimization
  of web page layout

25
Eye-Movement Studies

• Improving advertisements with eye-movement studies

26
Eye-Movement Studies
27
Eye-Movement Studies

• The technique of attentional landscapes (see
  previous slide) was used to visualize attention
  in ambiguous pictures.
• Subjects viewed ambiguous pictures that allowed
  two different interpretations A and B.
• They were asked to press and hold a particular
  button while perceiving interpretation A,
  and a different button for interpretation B.
• Afterwards, the fixations recorded during the
  perception of interpretations A and B were
  separated and separately visualized in the
  original image.

28
Looking at Visual Attention
29
Looking at Visual Attention
30
Eye Movement Studies

• Eye movements while watching (green), starting to
  imagine (black), and imagining a shape (red) -
  reading ones mind

31
Eye-Movement Studies

• Eye movements as
• indicators of cognitive
• processes (Yarbus)
• trace 1 examine at will
• trace 2 estimate wealth
• trace 3 estimate ages
• trace 4 guess previous activity
• trace 5 remember clothing
• trace 6 remember position
• trace 7 time since last visit

32
Eye Movement Studies
33
Eye Movement Studies
34
Eye Movement Studies
Gaze-contingent display in a visual search task
35
Eye Movement Studies
To be continued