Head-EyeK: Head-Eye Coordination and Control Learned in Virtual Reality.

Area of Science:

• Human-Computer Interaction
• Computer Graphics
• Psychology
• Robotics

Background:

• Human head-eye coordination is complex, influenced by physiology, psychology, and intent.
• Existing gaze models in psychology and graphics lack generalizability for realistic head-eye coordination.

Purpose of the Study:

• To develop and validate a novel head-eye coordinated Inverse Kinematic gaze model (Head-EyeK).
• To identify key factors influencing human head-eye coordination.
• To animate complex gaze behaviors credibly in virtual reality.

Main Methods:

• Developed an experimental protocol to collect head-eye motion data in Virtual Reality (VR) during sequential look-at tasks.
• Utilized collected data to identify factors influencing head-eye coordination.
• Introduced the Head-EyeK model, integrating identified insights into an Inverse Kinematic framework.

Main Results:

• Evaluated Head-EyeK's algorithmic parameters' impact on gaze behavior.
• Demonstrated quantitative superiority over prior art using ground-truth data.
• Showcased qualitative improvements via a perceptual study and credible animation of complex gaze behaviors.

Conclusions:

• Head-EyeK provides a significant advancement in modeling human head-eye coordination.
• The model credibly animates complex gaze behaviors, outperforming existing methods.
• This research contributes to more realistic human-like interactions in virtual environments.