EUROSPEECH 2003 - INTERSPEECH 2003
We show how to construct a channel-independent representation of speech that has propagated through a noisy reverberant channel. The method achieved greater channel-independence than cepstral mean normalization (CMN), and it was comparable to the combination of CMN and spectral subtraction (SS), despite the fact that no measurements of channel noise or reverberations were required (unlike SS).
Bibliographic reference. Levin, David N. (2003): "Blind normalization of speech from different channels", In EUROSPEECH-2003, 2929-2932.