[PDF] State-of-the-Art in Visual Attention Modeling | Semantic Scholar

The results showed that egocentric attention is very individual and differs from 2D image viewing conditions, partially due to binocular cues that enhance viewer’s perception, and suggested to employ specialized models for egocentric vision.

Three models of scanpaths, i.e. trajectories of free visual exploration, share a fundamental idea: the evolution of the mechanisms of visual attention has been guided by fundamental functional principles, and are suitable for real-time application.

• Highly Influenced

This thesis proposes three saliency models for in-depth investigation in modeling visual attention, including an unsupervised model using the bidirectional propagation method, a Convolutional Neural Networks based model by connecting the Dense Conditional Random Fields for multi-scale saliency refinement, and a convolutional neural network based model with cascade Conditional random Fields for joint model training.