Unsupervised Spectral Substraction for Noise-Robust ASR

This paper proposes a simple, computationally efficient \mbox{2-mixture} model approach to discriminate between speech and background noise at the magnitude spectrogram level. It is directly derived from observations on real data, and can be used in a fully unsupervised manner, with the EM algorithm. In this paper, the 2-mixture model is used in an ``Unsupervised Spectral Substraction'' scheme that can be applied as a pre-processing step for any acoustic feature extraction scheme, such as MFCCs or PLP. The goal is to improve noise-robustness of the acoustic features. Experimental results on both OGINumbers95 and Aurora~2 tasks yielded a major improvement on all noise conditions, while retaining a similar performance on clean conditions.