Height Judgment in VR

1
Height Judgment in VR

• The effect of scale

2
Introduction

• This project seeks to determine the significance
  of perspective imaging versus familiar cues when
  judging heights in a virtual environment.
• By rescaling familiar objects we believe we can
  have a larger effect on a subjects perception of
  height than the surrounding perspective visual
  cues.

3
Other Studies

• People in VR environments tend to underestimate
  distances (various)
• Minification influences spatial judgements in VR
  (Kuhl, Thompson, Creem-Regher)
• Absolute motion parallax weakly determines visual
  scale in real and virtual environments (Beall,
  Loomis, Philback, Fikes)
• The quality of computer graphics does not matter
  when judging distance in VR (Willemsen, Gooch,
  Creem-Regehr, Loomis, and Beall)
• The relative horizon height and perceived
  egocentric distances are significantly correlated
  (Franz, von der Heyde, Bulthoff)

4
Experimental Design
5
Pseudo-random Scale

• 3 Environments
• Plane
• Scale NA
• Hospital Chairs
• Scales 1x, 2x, 5x
• Campus Buildings
• Scales 0.5x, 1x, 5x

6
Pseudo-random Height

• Plane
• 0-100 ft
• Chairs
• 1x 0-10 ft
• 2x 0-20 ft
• 5x 0-50 ft
• Buildings
• 0.5x 0-50 ft
• 1x 0-100 ft
• 5x 0-500 ft

7
Results

• Percentage error
• Mean of ((Actual / judgment) / Actual) 100
• Mean percentage error
• Plane
• 10.26 ft
• 1x Chairs
• 0.57 ft
• 2x Chairs
• 18.13 ft
• 5x Chairs
• 64.4 ft
• 0.5x Campus
• -91.57 ft
• 1x Campus
• 0.93 ft
• 5x Campus
• 79.70 ft

8
Difference in Judgment vs. Scale
9
Error in Judgment vs. Scale
y -35.131x2 231.28x 198.43 y 31.142x
71.331
y 15.776x 14.264
10
Conclusion

• Subjects overestimated height when familiar
  objects appeared at a larger scale and
  underestimated height when familiar objects
  appeared at a smaller scale.
• However, studies of judgment at more scale levels
  should be conducted to determine exact trends