A dorsal pathway guided visual attention model

Abstract

Attention computational model is widely used in the embedded intelligent vision system to help it offload the processing effort. In this paper, we proposed a visual attention computational model based on the biological mechanism that the dorsal pathway will guide the ventral visual information process. The model involves two feature processing subsystems, one is the dorsal pathway feature processing subsystem and the other is the ventral pathway feature processing subsystem. Moreover, the dorsal pathway feature processing subsystem will generate a signal based on its processing result to modulate the information processing of the ventral pathway feature processing subsystem. The experiment results show that the proposed model outperforms the comparison models in four different test scenarios, which indicates that the proposed model may be more biologically plausible and can help the embedded intelligent vision system to find out the interested objects more accurately.