支配眼并非始终相同：视角、手性和3D中的眼动效果

研究背景与问题

• 作者发现了哪些问题或挑战？ Eye dominance, the subconscious preference for one eye over the other, is an important consideration in designing human-computer interfaces (HCI), especially for VR and AR systems. Previous studies in HCI assumed that eye dominance is static, but existing research in psychology suggests it can be dynamic and influenced by contextual factors such as viewing angle, handedness, and eye movement patterns. The lack of understanding of these dynamics limits the effectiveness of systems that depend on eye dominance for interaction accuracy and performance (e.g., foveated rendering, gaze-based interaction).

• 为什么这个问题很重要？ With the increasing adoption of immersive technologies like VR and AR, robust interaction methods that leverage eye dominance are critical to improving usability, rendering precision, and user experience. Assuming static eye dominance fails to account for user behavior variability, which can result in decreased performance and interaction fidelity.

• 研究动机与相关工作 Previous studies focused primarily on static targets, ignoring dynamic conditions such as moving or sequential targets, which are more common in AR/VR contexts. This study seeks to explore dynamic factors like random and sequenced saccades and sustained pursuit movements that influence eye dominance in VR environments, aiming to inform the design of adaptive systems capable of handling eye dominance fluctuations.

解决方案

• 作者提出了哪些方法或解决方案？ The authors designed three VR alignment tasks to study the dynamic nature of eye dominance by introducing different gaze behaviors:

  1. Task 1: Random saccades with pseudo-random horizontal target positions.
  2. Task 2: Sequenced saccades that presented targets in a sequence across a horizontal axis.
  3. Task 3: Pursuit movements where targets moved continuously across the horizon.

  They used a geometric measurement approach to determine eye dominance dynamically without relying on subjective self-reporting methods.

• 该解决方案的创新之处是什么？ The study innovatively adapts prior methods for eye dominance measurement to dynamic target behavior and avoids biases from traditional self-reporting techniques. By analyzing alignment tasks through precise geometric calculations, the authors remove reliance on subjective inputs and enhance the ability to analyze eye dominance shifts in real-time.

• 实施步骤是什么？使用了哪些关键技术？ The experimental setup was developed in VR with tools like Unity and SteamVR and utilized hardware (HTC VIVE Pro Eye HMD) combined with geometric measurement:

  1. Eye dominance detection: The distance between the dominant eye and cursor alignment was measured without subjective reporting.
  2. Controlled experimental conditions: Tasks were presented with randomized order, supported by rigorous geometric calculations based on the physical positions of the eyes and the VR cursor.
  3. Analysis: Binomial logistic regression was performed to evaluate the influence of variables like viewing angle, movement direction, and handedness.

研究成果

• 取得了哪些具体成果？

  1. Eye dominance shows variability depending on conditions, but shifts are less frequent than prior research suggested.
  2. Random saccades (Task 1) showed that neither viewing angle nor hand used significantly predicted eye dominance.
  3. Sequenced saccades (Task 2) revealed that movement direction influenced which eye was used for alignment, but the predictive capacity remained weak.
  4. Pursuit movements (Task 3) demonstrated a significant influence of hand used, target angle, and movement direction on eye dominance.
  5. Approximately two-thirds of participants showed consistent right-eye dominance, aligning with pre-experiment measurements.

• 与现有解决方案相比，它有哪些优势？ The authors developed a robust, objective geometric measurement framework for assessing eye dominance in dynamic conditions, addressing biases seen in traditional methods. Their approach provides detailed insights into eye dominance shifts across different scenarios and lays the foundation for adaptive designs that accommodate individual variability.

• 实验或评估结果是什么？ Switches in eye dominance occurred but were limited to specific conditions (e.g., centripetal movement during pursuit). Across all participants:

  • Right-eye dominance predominated but was influenced by factors like movement direction.
  • Task complexity (e.g., smooth pursuit movements) was likelier to elicit shifts.

• 局限性与未来方向

  1. Limitations:
     • The study focused only on horizontal viewing angles, potentially limiting the generalizability of findings.
     • Hand tremors or task repetition may have indirectly influenced participant behavior.
     • A relatively small participant pool may reduce statistical power for exploring nuanced subgroup differences.
  2. Future directions:
     • Extend research to vertical and diagonal viewing angles.
     • Explore additional influencing factors such as visual acuity, target size, and cognitive load.
     • Develop structural equation models (SEM) to study multivariate influences on eye dominance.
     • Design hands-free methods to test eye dominance using neural or gaze-based technologies.

总结

This study enhances our understanding of eye dominance and its dynamics in VR environments by introducing innovative measurement methods and dynamic alignment tasks. The findings challenge the assumption of static eye dominance and emphasize the importance of adaptive interfaces in AR/VR systems. Applications like foveated rendering, gaze-based selection, and immersive content placement can benefit from integrating these insights for personalized and adaptive user experiences. Future studies should address limitations by exploring further complexities of eye dominance behavior.