Georgia Tech developes 3D gaze forecasting for AR devices
TL;DR. Researchers at Georgia Tech developed a 3D gaze forecasting method for AR devices, allowing them to proactively render scenes before a user looks. - The new approach predicts user attention in 3D, unlike existing 2D gaze prediction research, which improves AR interaction. - This technology enables AR systems to anticipate user interaction, creating a more seamless and responsive experience. - Research used Meta's Aria Digital Twin dataset, which includes first-person video and detailed 3D environment reconstructions.
- Georgia Tech Ph.D. student Fiona Ryan developed 3D gaze forecasting for AR devices.
- This method predicts user attention in 3D environments, moving beyond existing 2D gaze prediction techniques.
- The research aims to allow AR systems to proactively render scenes and interactions, enhancing user experience.
- Ryan's work, conducted during an internship at Meta, utilized the Meta Aria Digital Twin dataset.