Real time voice active detection
- Detect default parameters during a silent period.
- Calculate the short-term energe(E) of one frame.
- Calculate the Spectral Flatness Measure(SFM) and the dominant frequency component(F) of the frame.
- Compare the feature of the frame with the default parameters to figure out whether the voice is active in this frame.
- Update the default parameters according to the new frame detected.
Moattar, Mohammad & Homayoonpoor, Mahdi. (2010). A simple but efficient real-time voice activity detection algorithm. European Signal Processing Conference.