Abstract
In order to find out a particular user under complex environments, the robot can make full use of the information of interaction. Especially in the case that the robot has no information about the user on ahead, the method of human recognition or speech recognition can't be used. The way of using efficient interaction to look for the user will be more effective. Let the robot try to talk to the user, and detect the changes (both visual and auditory information) of the environment when the user makes a response. By repeating this process several times, the user can be finally fixed. Furthermore, this method can handle any complex environment by making some special requirements to the people around the robot, e.g. to request noisy people to be quiet.
