A novel bare-handed manipulation technique for distant objects in virtual reality

Advancements in bare-hand tracking technology, featured in many modern virtual reality (VR) headsets, have enabled hand interaction for proximate 3D object manipulation. However, manipulating distant 3D objects in VR with bare hands remains a challenge and my work explores the potential of two-handed interaction in this context. Previous work highlighted the ease of using the non-dominant wrist as an interactive surface for interaction, by pointing at it with the dominant hand's index finger. My research expands this paradigm by extending the interactive zone to the entire non-dominant hand, so that the user can interact with their index finger on the non-dominant hand's fingers and palm to adeptly and precisely manipulate a distant 3D object's position and rotation. Through a user study, I compare my method with HOMER in terms of efficiency, accuracy, and usability. The results demonstrate a notable accuracy improvement (at least 75% better), especially in executing complex tasks, such as hanging a painting on a slanted wall, albeit at the expense of efficiency (about 47% slower) when compared to HOMER. While HOMER is faster, it also exhibits more variation in time, and at the average HOMER end time, my new technique is still substantially more accurate than HOMER. Further, my method received more favorable ratings on the System Usability Scale (SUS), underscoring the user-centric and intuitive design of my approach. Furthermore, the NASA Task Load Index (NASA-TLX) indicates that my technique not only improves task completion rate and user satisfaction but also reduces cognitive, physical, and temporal demands on users. In addition, I discuss insights into error prevention, the integration of constraints, and the facilitation of consistent interaction. My findings lay a solid foundation for the application of my new method in the dynamically advancing field of VR systems operated without controllers.