방향 추정용 로봇 인공 귀를 위한 반사파 제거 신호처리 기법

Sound source localization (SSL) is a listener’s capability to estimate the direction or position of detected sound. It may indicate the various methods to achieve SSL using the measurements of the acoustic signals from microphone arrays in acoustical engineering as well. This is one of the important techniques used for Human-Robot Interaction (HRI). Robot industry is demanding some specific type of auditory system based on artificial ear shaped after human external ear for more natural HRI. We propose a specific sound direction estimation method using a pair of artificial ears, each of which consists of a single ear pinna and two microphones. In this paper, we focus on cleansing method of reflected wave, one of the essential parts of sound direction estimation. We studied and analyzed the cleansing process. In result, we showed that performance of sound direction estimation is improved in spite of environmental noises such as reflected wave and so on.