Sixth European Conference on Speech Communication and Technology
This paper addresses the problem of robust text-independent speaker verification in the presence of missing (masked by noise) features. It presents and assesses several missing feature handling approaches. In these approaches, the speech enhancement and missing feature detection are based on the minimum mean-square error (MMSE) spectral amplitude estimator of Ephraim and Malah .
Full Paper (PDF) Gnu-Zipped Postscript
Bibliographic reference. El-Maliki, Mounir / Drygajlo, Andrzej (1999): "Missing features detection and handling for robust speaker verification", In EUROSPEECH'99, 975-978.