A Visual Attention Model Using Depth Information from the Point of Gaze