Japanese voice interface system with color image for controlling robot manipulators

In this paper, voice commands based on natural spoken language are used so that two or more same colored objects are tracked by a robot manipulator. More precisely, after receiving a voice command regarding the color information, the end-effector of the robot is controlled to approach a desired object out of many objects by visual feedback, in which the visual information is further applied to the human if a more correct motion of the robot is required. Thus, the present objective is to obtain a smoother cooperative system between human and robot, by coordinating voice and visual information.