Microphone Mini-array Based Speech Enhancement Using ISDS-MGSC Algorithm
Microphone Mini-array based speech enhancement is a challenging research subject since the aperture of microphone array is greatly limited. An efficient algorithm is presented in this paper for microphone mini-array based speech enhancement systems. The algorithm proposes an Improved Shared Distorted Signal (ISDS) method for Modified Generalized Sidelobe Canceller (MGSC). The idea is to apply ISDS to cancel the speech signal in the blocking subsystem of MGSC, which introduces noticeable speech improvement. In experiment an average improvement over several noise sources settings of ~14dB is achieved as compared to ~3dB improvement by MGSC.
Qingning Zeng Qinghua Liu Shan Ouyang Waleed Abdulla
Guilin University of Electronic Technology, Guilin University of Auckland, Auckland, New Zealand
国际会议
2011亚太信号与信息处理协会年度峰会(APSIPAASC 2011)
西安
英文
1-4
2011-10-18（万方平台首次上网日期，不代表论文的发表时间）