Gain function estimation for TDC-based subspace speech enhancement

In this paper, we present a gain function estimation method for time domain constrained (TDC) based subspace speech enhancement. The optimal Lagrange multiplier of the gain function will be estimated in accordance with signal to noise ratio (SNR) of the noisy speech. This gain function is placed between a Karhunen-Loeve transform (KLT) and an inverse KLT to construct an optimal estimator for speech enhancement. The superiority of the proposed gain function estimation method is demonstrated by experimental comparison with two classic approaches proposed by Ephraim and Van Trees.