A Novel Biologically Inspired Attention Mechanism for a Social Robot

摘要

In biological vision systems, the attention mechanism is responsible for selecting the relevant information from the sensed field of view. In robotics, this ability is specially useful because of the restrictions in computational resources which are necessary to simultaneously perform different tasks. An emerging area in robotics is developing social robots which are capable to navigate and to interact with humans and with their environment by perceiving the real world in a similar way that people do. In this proposal, we focus on the development of an object-based attention mechanism for a social robot. It consists of three main modules. The first one (preattentive stage) implements a concept of saliency based on "proto-objects." In the second stage (semiattentive), significant items according to the tasks to accomplish are identified and tracked. Finally, the attentive stage fixes the field of attention to the most salient object depending on the current task.