Cognitive Issues in VR Chapter 13 Wickens

1
Cognitive Issues in VRChapter 13Wickens Baker
2
Cognitive Issues

• Perceiving
• Understanding
• Short term
• Where am I , what do I see, where do I go, how?
• Long term
• What can I learn

3
5 Features of VE

• 3-D vs 2-D viewing
• 3D more realistic and similar to how we view the
  real world.
• Dynamic vs static display
• E.G.. Movie may provide better understanding than
  static pictures of same information.
• Closed-loop vs open-loop interaction
• Closed-loop person has control over the aspects
  of the world. Active vs passive

4
5 Features of VE

• Ego-referenced vs world reference
• Ego vs exocentric view points.
• Frame of reference.
• Ego reference considered immersed in environment.
• Multimodel interaction
• Our interactions and learning in the real world
  is based on information input into multiple
  channels.
• Vision, auditory, touch, kinesthetic feedback,
  gesture, etc...

5
5 Features of VE

• Five elements are not independent
• We dont know what must be included
• Research focuses on subsets of VE features
  because of difficulty of incorporating everything
• What you need will depend on the purpose of the
  environment

6
Uses of VE

• According to Wickens Baker, there are four
  categories of tasks that VE systems are designed
  to serve
• 1. On-line performance - manipulation in a remote
  environment. Issues include
• Closed loop performance
• Situation awareness
• Workload or cognitive effort

7
Uses of VE

• 2. Off-line training and rehearsal - practice in
  forgiving environment. Issues include
• Transfer of training
• How realistic must it be?
• 3. On line comprehension - obtaining
  understanding, comprehending, insight of
  information. Issues include
• How should the information be modeled?
• Will modeling choices affect understanding
  incorrectly?

8
Uses of VE

• Off-line learning and knowledge acquisition-
  transfer knowledge to be employed later is
  different form (e.g. Education)
• Does it improve understanding more than other
  methods.
• How can deeper knowledge be taught in this
  environment?

9
Tasks in Ves

• Human factors principles should be applied in the
  development of VE. Including analysis of user
  needs and tasks.
• Wickens baker discuss 6 tasks that may be
  performed in a VE.
• Search navigation
• Visual-motor coupling manipulation
• Perception and inspection learning

10
Tasks in Ves Search

• Search - looking for information in a 3D world,
  database, image, etc.
• Maps
• Different frames of reference
• Aids (cues) to help with search

11
Tasks in Ves Navigation

• Navigation
• Where am I, where do I go?
• Point to point vs smooth navigation
• Travel metaphor (flight, walking, logarithmic)
• Situation awareness
• Control devices and ability to navigate
• Stability of perceptual motor loop
• Perceptual biases

12
Perceptual Bias

• Humans use cues to distinguish locations and
  orientations of self and objects in their
  environments such as
• Regularly spaced texture, size, binocular cues,
  gradients, slant, optic flow, etc.. (Depth
  perception cues).
• These cues need to be included in the
  environment, but designed cues may not provide
  accurate perception of object location.
• Examples of two perceptual biases
• Virtual space effect
• 2D-3D effect

13
Perceptual Biases Cont.

• Virtual space effect - bias in perceived location
  of objects because of display magnification (
  wide field of view) or minification (narrow field
  of view).
• 2D-3D effect - perceived rotation of vectors
  towards a plane that is more closely parallel to
  the viewing plane. (Under estimation of slant of
  a surface, figure 13-1).

14
Perceptual Biases Cont.

• Research needs
• What are the perceptual biases?
• What designs or aids can we include to help
  users?
• What tasks are affected?

15
Tasks in VEVisual Motor Coupling

• The VE system can introduce artifacts that
  interrupt visual-motor coupling (e.G..Reaching,
  turning the head and seeing update of visual
  scene). Five issues are
• Gain
• Time delay
• Order
• Travel-view decoupling direction of motion
  decoupled from direction of gaze
• Field-of-view

16
Visual Motor Coupling Gain

• Gain setting on control input.
• Low gain, need to move the input a great deal
  before change occurs.
• High gain, small input, large change or
  displacement. (Can lead to instability of
  control).
• Appropriate gain settings will depend on input
  device and task.
• Example space ball rotation vs translation

17
Visual Motor Coupling Time Delay

• Time delays or lags in updating the VE.
• Flight simulation shows that 50 ms can impact
  performance.
• May also effect spatial sensory reflexes (VOR).
• To reduce time delays consider
• processing required due to complexity of images
• adaptive or progressive imagery (simplified
  views)
• predictive filtering of user input
• research is still needed

18
Visual Motor Coupling Control Order

• System state - Task Dependent
• zero order - position control
• First-order - constant rate of change
• use when navigation for movement to locations

19
Visual Motor CouplingTravel View Decoupling

• Direction of movement of travel, decoupled from
  direction of gaze.
• Difficult to accomplish, especially with system
  lags
• May cause disorientation
• Needs research

20
Visual Motor Coupling Field of View

• Field of view Vs viewing angle
• Field of view - how much of the field we can see
  at once (Instantaneous , without moving head).
• Viewing angle is width subtended by a display.
• Field of regard - with limited field of view can
  rotate our head and get larger field of regard.
• Real world viewing angle field of view
• VE can increase field of view as viewing angle
  is held constant (wide field of view,
  minification) or decrease field of view
  (magnification)

21
Visual Motor Coupling Field of View

• If use wider field of view
• see more information in the scene at once, may
  help with situation awareness
• distort perceived location of objects
• produced compelling sense of forward motion
• may result in greater mismatch between motion and
  vestibular inputs

22
Tasks in VE Manipulation

• Issues for manipulating objects are similar to
  those for Navigation and Visual Motor Coupling
• gain, time lag, control order, depiction of
  location and orientation, frame of reference
• also consider feedback (force or auditory)

23
Tasks in VE Perception and Inspection

• Features need for this task may be different than
  those needed for Navigation.
• What is needed is task dependent
• May not need large amount of visual realism

24
Tasks in VELearning

• Procedural training
• how much realism do we need?
• Active participation more important than realism
• Perceptual motor skill learning
• active control
• sensory feedback
• Spatial learning and navigational rehearsal
• ability to build mental maps
• Conceptual learning (understanding phenomenon)
• active exploration
• alternative knowledge representations

25
Effort and Workload

• The VE could increase rather than workload.
  Need to evaluate the design. If interacting in
  the world becomes the primary task that
  diminishes the purpose, it may not be worth it.
• Forcing more effort to understanding
  relationships can at times be beneficial to
  learning.