Probabilistic learning of task-specific visual attention

Despite a considerable amount of previous work on bottom-up saliency modeling for predicting human fixations over static and dynamic stimuli, few studies have thus far attempted to model top-down and task-driven influences of visual attention. Here, taking advantage of the sequential nature of real-world tasks, we propose a unified Bayesian approach for modeling task-driven visual attention. Several sources of information, including global context of a scene, previous attended locations, and previous motor actions, are integrated over time to predict the next attended location. Recording eye movements while subjects engage in 5 contemporary 2D and 3D video games, as modest counterparts of everyday tasks, we show that our approach is able to predict human attention and gaze better than the state-of-the-art, with a large margin (about 15% increase in prediction accuracy). The advantage of our approach is that it is automatic and applicable to arbitrary visual tasks.