We have built a robot system that can acquire the ability for sound source localization by self-organization through repetition of motion and sensing. This robot consists of two microphones (auditory: ears), a video camera (visual: eye), and a rotation stand (motion: neck). This system can learn without any external explicit supervision.
