Impossible spaces: maximizing natural walking in virtual environments with self-overlapping architecture.

Area of Science:

• Virtual Reality
• Human-Computer Interaction
• Computer Graphics

Background:

• Virtual environments (VEs) are limited by physical workspace size for natural locomotion.
• Maximizing exploration in VEs requires overcoming physical space constraints.

Purpose of the Study:

• Introduce and evaluate "impossible spaces" as a design mechanic to expand virtual environment size.
• Investigate user perception and experience of impossible spaces with natural locomotion.

Main Methods:

• Developed impossible spaces using self-overlapping architectural layouts to compress large VEs into smaller physical areas.
• Conducted two user studies to assess detection thresholds and perceptual effects of space compression.
• Combined redirection techniques with impossible spaces in an outdoor VE scenario.

Main Results:

• Users detected impossible spaces at 56% overlap in small rooms and 31% in larger rooms filling the workspace.
• Distance perception remained accurate even with noticeable overlap.
• Exploratory analysis indicated enhanced illusion with naive users in chained impossible spaces.

Conclusions:

• Impossible spaces effectively expand virtual environments for natural locomotion within limited physical areas.
• User perception of space and distance is robust to significant virtual space compression.
• Impossible spaces enhance the immersive illusion, particularly for users unaware of the manipulation.