Robust robot's attention for human based on the multi-modal sensor and robot behavior

Abstract

In this paper, we propose the robust robot's attention for human based on the multi-modal sensor and robot behavior. All of the robot components for attention are operating on the intelligent robot software architecture. Human search collect the human information from the sensing information for vision and voice in various illumination change and dynamic environments. And human tracker follows the face trajectory with efficiency and safety. Unlike common belief, the biggest obstacle and competitive factor in robotics is expected to be human robot interaction. Since robot intelligence is not yet at a practical level, the creation of a general interaction manager using an intelligence system will not be realized for some time. Instead, it focuses on one-way speaking and expressing based on emotion to human. Experimental results show that the proposed scheme works successfully in real environments. ？2010 IEEE.