Real-Time Robust Voice Activity Detection Using the Upper Envelope Weighted Entropy Measure and the Dual-Rate Adaptive Nonlinear Filter

Real-Time Robust Voice Activity Detection Using the Upper Envelope Weighted Entropy Measure and the Dual-Rate Adaptive Nonlinear Filter

Voice activity detection (VAD) is a vital process in voice communication systems to avoid unnecessary coding and transmission of noise. Most of the existing VAD algorithms continue to suffer high false alarm rates and low sensitivity when the signal-to-noise ratio (SNR) is low, at 0 dB and below. Ot...

Full description

Bibliographic Details
Main Authors: Wei Qing Ong, Alan Wee Chiat Tan, V. Vijayakumar Vengadasalam, Cheah Heng Tan, Thean Hai Ooi
Format: Article
Language:English
Published: MDPI AG 2017-10-01
Series:Entropy
Subjects:
voice activity detector (VAD)
gammatone filter
asymmetric nonlinear filter
weight factor
entropy
dual-rate adaptive nonlinear filter
Online Access:https://www.mdpi.com/1099-4300/19/11/487

Internet

https://www.mdpi.com/1099-4300/19/11/487

Similar Items