Method and background estimator for voice activity detection

摘要

The present invention relates to a method and a background estimator in voice activity detector for updating a background noise estimate for an input signal. The input signal for a current frame is received and it is determined whether the current frame of the input signal comprises non-noise. Further, an additional determination is performed whether the current frame of the non-noise input comprises noise by analyzing characteristics at least related to correlation and energy level of the input signal, and background noise estimate is updated if it is determined that the current frame comprises noise.

主权项

1. A method for updating a background noise estimate for an input signal in a background estimator in a Voice Activity Detector, comprising:
receiving the input signal for a current frame; determining whether the current frame of the input signal comprises non-noise; determining whether the current frame of the input signal comprises noise by analyzing characteristics related to at least one of a correlation and an energy level of the input signal, after determining that the current frame comprises non-noise; and updating the background noise estimate in response to determining that the current frame comprises noise, wherein determining whether the current frame of the input signal comprises noise further comprises at least one of: detecting correlation and counting a number of frames from a frame last indicating a correlation event, in response to determining that a first difference between an energy level of the input signal and a smooth minimum energy level is within a first range and that a second difference between a noise level of the current frame and the smooth minimum energy level is within a second range.