Gaze Beyond the Frame: Forecasting Egocentric 3D Visual Span
This addresses a fundamental gap in egocentric scene understanding for AR/VR and assistive technologies, though it is incremental by extending existing 2D methods to 3D.
The paper tackles the problem of forecasting where a person's visual perception will focus next in a 3D environment, proposing EgoSpanLift to transform 2D gaze prediction into 3D scenes and achieving superior performance over baselines in egocentric 2D gaze anticipation and 3D localization.
People continuously perceive and interact with their surroundings based on underlying intentions that drive their exploration and behaviors. While research in egocentric user and scene understanding has focused primarily on motion and contact-based interaction, forecasting human visual perception itself remains less explored despite its fundamental role in guiding human actions and its implications for AR/VR and assistive technologies. We address the challenge of egocentric 3D visual span forecasting, predicting where a person's visual perception will focus next within their three-dimensional environment. To this end, we propose EgoSpanLift, a novel method that transforms egocentric visual span forecasting from 2D image planes to 3D scenes. EgoSpanLift converts SLAM-derived keypoints into gaze-compatible geometry and extracts volumetric visual span regions. We further combine EgoSpanLift with 3D U-Net and unidirectional transformers, enabling spatio-temporal fusion to efficiently predict future visual span in the 3D grid. In addition, we curate a comprehensive benchmark from raw egocentric multisensory data, creating a testbed with 364.6K samples for 3D visual span forecasting. Our approach outperforms competitive baselines for egocentric 2D gaze anticipation and 3D localization while achieving comparable results even when projected back onto 2D image planes without additional 2D-specific training.