Nonnegative matrix factorization based noise robust speaker verification

摘要

The performance of speaker verification system (SVS) declines dramatically in noisy environments. To suppress the adverse impact of the noise on SVS, this paper investigates employing the nonnegative matrix factorization (NMF) technique to reconstruct the speech based on the pre-trained speech basis matrix (SBM) and noise basis matrix (NBM). The contribution of this research lies in utilizing the time correlation of the speech signal to obtain a more appropriate SBM. An enhanced NMF-based speech enhancement algorithm (ENMF-SE) is derived. Accordingly, the robust SVS based on ENMF-SE (ENMF-SE-SVS) is constructed and evaluated by intensive experiments with a public speech database. Experimental results show that the proposed ENMF-SE-SVS provides up relative improvement EER compared with the traditional NMF-SE based SVS algorithm under different SNR noise conditions.