HAGI: Head-Assisted Gaze Imputation for Mobile Eye Trackers

Imputed Gaze Data Gaze Data with Missing ValuesFigure 1: Missing data is inevitable in mobile eye trackers.HAGI is a novel multi-modal approach for gaze data imputation that exploits the close coordination between eye and head movements.Input to our method is gaze data with missing values and time-aligned head movements captured using sensors readily available in mobile eye trackers (Left).HAGI uses a novel head-conditional diffusion model to impute the missing gaze data (Right) with lower mean angular error and is more realistic than previous methods.