In this paper, we propose new approaches to speech enhancement based on soft decision. In order to enhance the statistical reliability in estimating speech activity, we introduce the concept of a global speech absence probability (GSAP). First, we compute the conventional speech absence probability (SAP) and then modify it according to the newly proposed GSAP. Moreover, for improving the performance of the SAP’s at voice tails (transition periods from speech to silence), we revise the SAP’s using a hang-over scheme based on hidden Markov model (HMM).

In addition, we suggest a robust noise update algorithm in which the noise power is estimated not only in the periods of speech absence but also during speech activity by noise and speech spectrum estimation based on soft decision. Also, for improving the SAP determination and noise update routine we present a new signal to noise ratio (SNR) concept which is called the predicted SNR in this paper. The prediced SNR is defined by the ratio between estimated speech and noise spectrum makes a further improvement the discrete cosine transform (DCT). Results from the test show that the proposed algorithm which is called the speech enhancement based on soft decision (SESD) yields better performance than the conventional methods.