Ninja Hands: Using Many Hands to Improve Target Selection in VR

Title of the Paper

Ninja Hands: Using Many Hands to Improve Target Selection in VR

Paper Information

• Field of Study: Target selection and interaction techniques in Virtual Reality (VR)
• Keywords: Virtual Reality, multi-hand systems, user study, target selection, virtual hands

Research Background and Problem

• Problem or Challenge:

  • In VR environments, users typically select and manipulate objects using virtual hands, but target selection efficiency decreases for objects beyond arm's reach.
  • Existing techniques for addressing long-distance interaction (e.g., ray casting or virtual hand extension) have limitations in precision and convenience.
  • Whether mapping multiple virtual hands to a single physical hand can improve target selection remains unclear.

• Significance:

  • As VR technology rapidly evolves, improving interaction efficiency and user experience is crucial for the widespread adoption of VR applications, especially in scenarios with large target distribution areas.

• Research Motivation and Related Work:

  • Inspired by "Ninja cursors," a method that significantly reduces target selection time by using multiple cursors.
  • Previous studies on virtual hand extensions (e.g., adding fingers or a third arm) have primarily focused on body ownership and user experience, lacking evaluations of task performance.
  • There is a need to explore the effectiveness of similar multi-hand systems in VR, balancing efficiency improvements with potential conflicts in user experience.

Solution

• Proposed Solution:

  • Introduced a target selection technique called "Ninja Hands," which uses a single physical hand to simultaneously control multiple virtual hands distributed in the environment.

• Innovations:

  • Applied multiple virtual hands in a three-dimensional VR space and developed mechanisms for determining the number of virtual hands, their arrangement, motion mapping, and hand selection.
  • Explored the trade-off between efficiency improvement and physical movement reduction in target selection, offering a novel approach for large-scale VR target selection.

• Implementation Steps and Key Techniques:

  • Number and Arrangement of Virtual Hands:
    • Virtual hands can be arranged in one-dimensional, two-dimensional grids, or three-dimensional cubes based on target distribution characteristics.
    • The distribution of hands reduces the distance between targets and the nearest hand, theoretically improving efficiency.
  • Mapping Mechanism for Virtual Hands:
    • Used constant multiplier mapping ratios to control virtual hand movements, avoiding the precision loss associated with traditional methods like nonlinear mapping.
  • Hand Selection Algorithm:
    • When multiple virtual hands simultaneously touch a target, a "queue algorithm" is used to ensure only one hand is activated at a time.

Research Results

• Main Findings:

  • First Study:
    • In a small target space (2.5m × 2.5m × 2.5m), configurations with 4 and 8 hands significantly reduced target selection time (by 5.62% and 9.46%, respectively) and physical movement (by 30.29% and 53.94%, respectively).
    • Users rated the multi-hand system's satisfaction level as comparable to the single-hand system.
  • Second Study:
    • In a larger space (10m × 5m × 10m), the performance of 8, 27, and 64 hands was compared.
    • Increasing the number of hands did not further reduce target selection time but significantly reduced physical movement distance.
    • Increasing the number of hands led to a linear increase in decision-making time, as users spent more time selecting which hand to use rather than moving.

• Advantages Over Existing Solutions:

  • Significantly reduced users' physical movement, especially in scenarios with high target density or wide target distribution.
  • Provided a natural hand-centered interaction experience, avoiding abstract selection mechanisms (e.g., ray casting or complex extended hand controls).

• Limitations and Future Directions:

  • Limitations:
    • The study only explored mapping for one physical hand and did not investigate the potential of dual-hand multi-hand interactions.
    • The impact of the system on body ownership perception and user presence has not been studied.
    • Further research is needed to evaluate performance improvements in object manipulation and movement tasks.
  • Future Directions:
    • Investigate the application of multi-hand systems in tasks involving weight perception and group object manipulation.
    • Combine current technologies (e.g., ray selection or dynamic cluster generation algorithms) to optimize the distribution and selection mechanisms of multi-hand systems.
    • Extend to full-body interactions, studying how to enhance user experience in richer virtual scenarios.