The RayHand Navigation: A Virtual Navigation Method with Relative Position between Hand and Gaze-Ray

Title of the Paper

RayHand Navigation: A Study on Virtual Navigation Methods Based on the Relative Position of Hand and Gaze Rays

Paper Information

• Field of Study: Human-Computer Interaction, Virtual Reality (VR) Navigation
• Keywords: Virtual Reality, Navigation, Gaze Ray, Hand Interaction, User Experience Understanding, Learning Effect
• Published in: CHI '24
• Publication Year: 2024

Research Background and Issues

Background

Virtual reality navigation has been studied for decades, with its primary tasks being the control of travel speed and direction. Common navigation methods are categorized into walking, steering, selection, and manipulation navigation. Among these, steering techniques provide a rich navigation experience and efficiently reach destinations. However, many existing methods rely on a single input mode (e.g., gaze, hand, or body torso), each with the following limitations:

• Gaze Navigation: Suffers from the "Midas touch" problem (i.e., triggering unintended actions by mere eye gaze).
• Hand Navigation: Typically requires additional hardware for speed control and may interfere with hand manipulation of objects.
• Torso Navigation: Requires extra tracking devices and causes physical fatigue.

Issues

Existing navigation techniques either fail to effectively decouple users' "looking around" and "moving" behaviors or have limitations in terms of user experience, usability, and learning efficacy. This study aims to address these issues and explore whether combining gaze and hand interaction can lead to a better navigation method.

Motivation

Building on the challenges of gaze and hand interaction, this research seeks to design an innovative virtual reality navigation technique that avoids the "Midas touch" problem while eliminating the need for additional hardware. The goal is to provide a more natural and low-effort user experience. The study also evaluates this method against existing approaches (e.g., head-hand navigation and torso tilt navigation) and observes its learning effects.

Solution

Technical Design

A navigation technique named RayHand Navigation is proposed, which combines gaze rays and hand interaction to achieve dynamic control of speed and direction. Key features include:

1. Initialization of Gaze Ray Direction: Users can quickly mark the starting direction of navigation using their gaze.
2. Hand-Based Control: After navigation is initiated, users control navigation speed and direction through hand movements relative to the gaze ray.
   • Hand forward and backward movement controls speed (closer to the head slows down, farther from the head speeds up).
   • Hand left and right movement controls direction.
3. Decoupling Gaze and Navigation: Once navigation is initiated, gaze no longer controls direction, allowing users to freely look around.

Innovations

1. Combination of Gaze and Hand Interaction: For the first time, gaze and hand gestures are simultaneously utilized for virtual navigation.
2. Visual Feedback: State changes are intuitively indicated through ray color changes (green for activation, purple for non-navigation state).
3. Elimination of Additional Hardware Requirements: Navigation is achieved solely through natural hand movements without external control devices.

Implementation Steps

1. Navigation Activation and Termination:
   • Navigation is triggered by placing the hand on the gaze ray for 0.3 seconds to avoid accidental activation.
   • Navigation stops when the gaze ray color reverts to purple.
2. Speed Control:
   • Speed is calculated based on the distance between the hand and the head (minimum speed set at 1.44 m/s, maximum speed at 6 m/s).
3. Direction Control:
   • Direction is adjusted based on the lateral offset of the hand relative to the gaze ray.
4. Experiment Design:
   • Comparison of RayHand Navigation with head-hand (HH) navigation and torso tilt (TL) navigation.

Research Results

Experimental Findings

• User Experience:
  • RayHand and HH navigation outperformed TL navigation in terms of usability, task load, psychological demand, and overall user experience quality.
  • RayHand navigation provided a higher level of enjoyment, especially in small interaction spaces.
• Navigation Performance:
  • Head-Hand Navigation (HH) was more precise in direction control, resulting in shorter paths.
  • RayHand Navigation required less physical movement, making it easier to stay within safe boundaries.
• Learning Effect:
  • Across two rounds of experiments, RayHand navigation demonstrated significant learning effects, with reduced psychological burden and task load over time.

Advantages

1. Low Physical Demand: RayHand navigation minimizes the need for large-scale body movements, reducing physical fatigue.
2. Enhanced Experience: The combination of gaze and hand interaction provides users with a more natural navigation method while allowing free exploration.
3. Mitigation of Limitations: Effectively addresses the "Midas touch" problem and elegantly handles independent control of speed and direction.

Limitations and Future Directions

• Limitations:
  1. Current RayHand implementation has minor issues, such as gaze ray obstructing the view.
  2. Existing experiments have not included comparisons with "body-based walking" techniques or more complex navigation scenarios.
• Future Directions:
  1. Extend the method to three-dimensional navigation, supporting vertical interactions.
  2. Incorporate adaptive speed control to dynamically adjust navigation speed based on the distance to the target object.
  3. Explore applications of RayHand in more complex task scenarios (e.g., object manipulation and interaction).

In summary, RayHand Navigation demonstrates significant potential in combining visual and gesture interactions, offering valuable innovations for virtual reality navigation technologies.