Enhancement and segregation specific person's speech based on human auditory scene analysis

• Main Authors: Wei-Che Wu, 吳瑋哲
• Other Authors: Shuenn-Tsong Young
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/39372178856500478948

碩士 === 國立陽明大學 === 醫學工程研究所 === 95 === There are several strategies to separate speech from noise. However,effective methods to solve multi speech’s sources problem remain elusive. The well-studied Cocktail party problem is to analyze and deal with multi speech’s sources signal. Here, we focus on a specific speech source. By sing speaker recognition and speech recognition techniques extracting speakers’speech characteristics coupled with human auditory scene analysis, the speakers’speech characteristics can be determined more precisely and distinguished from one another. These observed differences can then be used for signal segregation and enhancement for specific speakers. Therefore, the users like children can enhance the receiving signal intensity form specific speech’s sources as their parents or teacher.