Output details

Visual Attention in Objective Image Quality Assessment: Based on Eye-Tracking Data

<23> This article investigates whether and to what extent the addition of saliency can be beneficial to objective image quality prediction. The research literature has, to date, focused on the extension of an image quality metric with a computational saliency model. Instead, we conducted eye-tracking experiments to obtain the “ground truth” data of visual attention, making the results independent of the reliability of a computational attention model. The findings are highly beneficial for the development of attention-based objective metrics for real-time image quality assessment.