Design And Implementation Of Robust Speaker Change Detection Mechanism

The main task of Speaker Change Detection mechanism is to automatically segment an audio signal containing multiple speakers so that,ideally,each voice segment contains only one speaker.It synthesizes many techniques such as speech signal processing and pattern recognition.It has a wide range of applications in speaker recognition,speaker dynamic monitoring,and audio information retrieval,which has become one of the hot topics in the field of speech signal processing in recent years.The main work of this paper is as follows:(1)Design and implement the basic Speaker Change Detection mechanism,complete the necessary tests and analysis;(2)Design and implement Overlapped Speech Detection mechanism based on GBDT;(3)Design and implement Voice Activity Detection mechanism based on Deep Learning,verify its effectiveness by performing tests on audios in various scenes,apply it on Speaker Change Detection mechanism and obtain some effects;(4)Design and implement Speech Enhancement mechanism based on Deep Learning,which guarantees the performance of Speaker Change Detection mechanism to some extent;...