Past Research
A computational model of the temporal dynamics of human vision
Developing an accurate computational model of human visual attention has been a long-standing challenge. Such a model may allow any system to select only relevant information from a complex and cluttered visual input.
Itti et al (1998) have implemented a preliminary model of the human visual system. The model uses primary features such as intensity, color, orientation, flicker, and motion to create a saliency map that represents the relevance of visual attention. However, this model does not take into account the temporal dynamics of the human visual system, which play a crucial role in enabling humans to select relavent information from visual scenes.
We have developed a new computational model of human vision that extends to the aforementioned work by incorporating two temporal dynamic features of the human visual system:
1) Instantaneous saliency depletion with gradual recovery, which simulates the ''Inhibition of Return'' effect (Posnet and Cohen (1984)).
2) Gradual saliency depletion with instantaneous recovery, which is derived from the ''Neural Adaptation'' theorem (Hartline (1940)).
Results show that the proposed algorithm substantially outperformed previous algorithms when only gradual depletion was incorporated, and instantaneous depletion improved the performance in some cases.
[Demo video ]