Balancing Accuracy and Embodiment: A Hybrid Perspective for Complex Visuomotor Tasks in VR

Paper Title

Balancing Accuracy and Embodiment: A Hybrid Perspective for Complex Visuomotor Tasks in VR

Publication Info

• Topic area: Virtual reality perspectives for motor control and user experience.
• Keywords: Virtual reality, embodiment, visuomotor tasks, perspective continuum, hybrid view, user experience, postural control, motor performance, balance training, HCI.

Background and Problem

• Problem / challenge: Traditional first-person perspectives (1PP) in VR are assumed to maximize embodiment and motor performance, but they can lead to occlusion and reduced spatial awareness. Conversely, third-person perspectives (3PP) enhance spatial awareness but reduce embodiment. The trade-offs between these perspectives in complex motor tasks remain underexplored.
• Significance: Understanding the interplay between embodiment, performance, and user preference is critical for designing effective VR systems for training, rehabilitation, and exergames.
• Motivation and related work: Previous research has shown that 1PP enhances embodiment while 3PP improves spatial awareness. However, the impact of hybrid perspectives and their utility in balancing these trade-offs for complex tasks, such as balance and motor control, is not well understood. This study builds on the perspective continuum model to address this gap.

Solution

• Proposed approach: Introduction of a novel Hybrid perspective that combines third-person spatial awareness with first-person limb feedback to optimize performance and embodiment in a VR balance game.
• Novelty:
  1. Systematic comparison of four perspectives (First, Ghost, Third, Hybrid) along the perspective continuum in a physically demanding VR task.
  2. Empirical evidence showing a mismatch between embodiment and user preference, with users favoring utility over immersion.
  3. Demonstration of the Hybrid perspective as a practical compromise for balancing performance and perception.
• Procedure and key techniques:
  • Conducted a within-subjects user study (N=20) using a VR balance game inspired by "Brain Wall."
  • Compared four perspectives under two difficulty levels (easy: 2-collectibles; hard: 3-collectibles).
  • Measured objective performance (success rate, wall touches, step-offs, task duration), balance metrics (oscillations, body lean), and subjective experience (presence, embodiment, workload, enjoyment).
  • Used a counterbalanced design and statistical analysis to evaluate effects.

Results

• Concrete findings:
  • Hybrid perspective improved postural stability (reduced head and waist oscillations) in the easy condition but showed no advantage in the hard condition.
  • Task duration was longer in Third and Hybrid perspectives under high difficulty, indicating a speed-accuracy trade-off.
  • First and Ghost perspectives maximized embodiment (Self-Location, Ownership) and presence but were less preferred.
  • No significant differences in error rates (success, wall touches, step-offs) across perspectives.
• Advantage over baselines:
  • Hybrid perspective offered a balance between spatial awareness and limb feedback, outperforming First and Ghost in user preference despite lower embodiment scores.
  • Third perspective consistently resulted in a more upright and stable posture.
• Experiments / evaluation:
  • Conducted 16 trials per participant (4 perspectives × 4 trials) with a counterbalanced order.
  • Used motion tracking for balance metrics and standardized questionnaires for subjective measures.
  • Participants ranked Hybrid as the most preferred perspective, followed by Third.
• Limitations and future work:
  • Small, homogeneous sample (N=20) limits generalizability; future studies should include clinical populations.
  • Results are task-specific; further research is needed across diverse VR tasks.
  • Only four discrete perspectives were tested; future work should explore the full perspective continuum.
  • Short exposure duration (12 minutes) limits conclusions about long-term adaptation.

Summary

This study investigated the trade-offs between embodiment, performance, and user preference in VR through a balance game comparing four visual perspectives. The novel Hybrid perspective, combining third-person spatial awareness with first-person limb feedback, emerged as the most preferred despite lower embodiment scores. Results highlight the importance of task-specific perspective design, showing that users prioritize utility over immersion in complex motor tasks. These findings have implications for VR training and rehabilitation, emphasizing the need for flexible, user-adaptive perspectives rather than a one-size-fits-all approach.