CNN-based estimation of gaze distance in virtual reality using eye tracking and depth data

Dataset to train models for predicting gaze distance, acquired during a VR experiment. The dataset consists of eye tracking data for 41 participants in an indoor and an outdoor VR setting, with corresponding depth and distance maps at 10deg around the measured gaze direction. The depth maps contain the scene depth with respect to the camera plane, while the distance maps contain the distance relative to the viewer. For more information please see the corresponding publication.