Voice activity detection using Laplacian model and UMP test
Keun Won Jang, Dong Kook Kim, Joon‐Hyuk Chang · Annual Conference on Computers · 2007
This paper presents a new voice activity detection (VAD) method using the Laplacian distribution and a uniformly most powerful (UMP) test. The UMP test is employed to derive the new decision rule based on likelihood ratio test (LRT). The proposed method provide the decision rule by comparing the sum of magnitude of real and imaginary parts of the noisy spectral component to the adaptive threshold estimated from the noise statistics. Experimental results show that the proposed VAD algorithm based on the Laplacian and the UMP test outperforms the conventional scheme at a low SNR.