Gaze behavior and visual attention model when turning in virtual environments