Speech denoising using nonnegative matrix factorization with priors

We present a technique for denoising speech using nonnegative matrix factorization (NMF) in combination with statistical speech and noise models. We compare our new technique to standard NMF and to a state-of-the-art Wiener filter implementation and show improvements in speech quality across a range of interfering noise types.

• Type: Conference
• Year: 2008
• Where: ICASSP
• Authors: Kevin W. Wilson, Bhiksha Raj, Paris Smaragdis, Ajay Divakaran
• Keywords: ICASSP 2008, Nonnegative Matrix Factorization, Signal Processing, Standard Nmf, State-of-the-art Wiener filter