Audio-Visual-Olfactory Resource Allocation for Tri-modal Virtual Environments.

Area of Science:

• Human-Computer Interaction
• Virtual Reality
• Sensory Perception

Background:

• Virtual environments (VEs) offer controlled simulations for training and entertainment.
• Realistic VEs necessitate multi-sensory, physically accurate stimuli.
• Simulating the full human sensory system (HSS) is computationally intensive.

Purpose of the Study:

• To propose a resource distribution scheme for optimizing perceptual experience in VEs within computational budgets.
• To investigate resource balancing for multi-modal scenarios involving aural, visual, and olfactory stimuli.
• To develop a quality prediction model for sensory stimuli in VEs.

Main Methods:

• Conducted three experimental studies to identify olfactory computational boundaries and user preferences.
• Recruited 25 participants to assess optimal visual, acoustic, and olfactory stimulus quality across 5 budget levels.
• Developed and validated a quality prediction model using collected data.

Main Results:

• Users initially prioritize visual quality over other sensory stimuli.
• Increased computational budgets lead to a preference for balanced resource distribution.
• A growing preference for olfactory stimuli inclusion was observed with larger budgets.

Conclusions:

• A novel resource distribution scheme can enhance perceptual experience in multi-modal VEs.
• User preferences for sensory input evolve with available computational resources.
• The proposed quality prediction model accurately forecasts user satisfaction across different budget allocations.