with various levels of background noise and acoustic signals in the voiced and unvoiced speech to the detection system and method for It is provided. The system (Figure 3) is to receive acoustic signals at two microphones (mic1, MIC2), two microphones (mic1, MIC2), to generate a difference between the acoustic parameters from the signal received by each. Difference parameter represents the relative difference between the signal portion of the received acoustic signal gain. System when the difference parameters exceed a first limit of the acoustic signal is classified as unvoiced speech information, the system when the difference parameter is greater than a second limit classifies the acoustic signal information to the voiced speech. Furthermore, the embodiment of the system according to the invention for receiving a non-physiological information to aid the voiced speech classification - and a sound sensor (20).
展开▼
