关键词: 谱减法, 语音增强, 非平稳噪声, Bark子带

Abstract: The traditional spectral subtraction can’t suppress the non-stationary noise in the realistic voice communication effectively.In order to improve the performance in eliminating the non-stationary noise，a noise estimation algorithm is proposed，which divides the power spectrum of the noisy speech into several subbands according to the Bark frequencies and tracks the noise by the improved minima controlled recursive averaging method in every individual band.The noise estimation approach can estimate the noise power spectrum more accurately while reducing the computation.Then the spectral subtraction method using the noise estimation approach is presented and compared with the traditional spectral subtraction.The result shows that the proposed method can suppress the noise better while reducing the musical noise more efficiently.

Key words: spectral subtraction, speech enhancement, non-stationary noise, Bark bands