OctoPocus in VR: Using a Dynamic Guide for 3D Mid-Air Gestures in Virtual Reality.

Area of Science:

• Human-Computer Interaction
• Virtual Reality
• Gesture Recognition

Background:

• Traditional 2D gesture guides like OctoPocus aid learning and recall.
• Adapting gesture guidance systems to 3D mid-air interactions in Virtual Reality (VR) presents unique challenges.
• Effective guidance is crucial for user proficiency in complex VR environments.

Purpose of the Study:

• To adapt the OctoPocus dynamic guide for 3D mid-air gestures in Virtual Reality.
• To evaluate the effectiveness of the adapted OctoPocus system compared to a VR crib-sheet for learning and performing gestures.
• To assess user experience and subjective performance in VR gesture learning.

Main Methods:

• Adapted the OctoPocus system for 3D mid-air gestures using an optimization-based recognizer.
• Introduced an exploration mode to visualize the complexity of 3D gesture guides.
• Conducted a comparative experiment replicating the original protocol, pitting OctoPocus in VR against a VR crib-sheet.

Main Results:

• OctoPocus in VR resulted in 1.8s faster gesture execution and 13.8% higher accuracy during training compared to a crib-sheet.
• Participants using OctoPocus remembered a comparable number of gestures despite a 0.9s increase in reaction time.
• Subjective feedback indicated 75% found OctoPocus easier to learn and 83% found it more accurate.

Conclusions:

• The adaptation of OctoPocus for 3D mid-air gestures in VR is feasible and beneficial.
• OctoPocus enhances VR gesture learning by improving execution speed and accuracy.
• This dynamic guidance approach offers a more effective method for mastering complex 3D gestures in virtual environments.