Abstract
The purpose of voice activity detection (VAD) is to classify voice and noise in an audio file. According to the fact that the spectral distribution of voice and noise segments in an audio file is different, a new VAD method is proposed. First, three-layer wavelet decomposition is executed to extract the multi-resolution spectrum of the audio signal. Then, the multivariate normal probability density function (MVNPDF) is calculated for each level of wavelet transform coefficients in the multi-resolution spectrum to obtain the feature - DWT-MVNPDF. Finally, based on the DWT-MVNPDF feature, dual-threshold method is used for VAD. The experiment results show that the proposed method is superior to the energy entropy ratio method (EER) and other algorithms at different SNR environments.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
