TDSal: Task-Based Top-Down Saliency Prediction Model

Can Mizrakli, Tolga K. Capin

Visual saliency aims to predict the regions of an image most likely to attract human visual attention. While most saliency models assume free-viewing conditions, human attention is…