A unified approach to speech enhancement and voice activity detection

In this work, a unied system for voice activity detection (VAD) and speech enhancement is proposed. In the proposed system, there is mutual exchange of information between VAD and speech enhancement blocks. A new, robust and low complexity VAD algorithm is implemented for the VAD block of the unied system. The newly proposed VAD algorithm uses a periodicity measure and an energy measure obtained from spectral energy distribution and spectral energy dierence of the input speech data. For the speech enhancement block, the Modied Wiener Filtering (MWF) algorithm is utilized. It has been shown that the utilization of information exchange between the VAD and MWF algorithms in the unied system increases the performance of both algorithms and the proposed unied system improves the robustness of a speech recognition system signicantly. Both of the enhanced algorithms are non-iterative. Therefore, the proposed unied system is computationally attractive for real-time applications.