An analysis of cepstral-time feature matrices for noise and channel robust speech recognition

Ben Milner, Saeed V. Vaseghi

Research output: Chapter in Book/Report/Conference proceedingConference contribution


This paper presents an analysis of the cepstral-time matrix. The coefficients of the cepstral-time matrix are found to be similar to the standard cepstral vector with differential features augmented on. It is also shown that the cepstral-time matrix is inherently robust to convolutional channel distortion. Spectral-subtraction, Wiener filtering and model combination are extended into two-dimensions where improved noise robustness is achieved. Experimental results using the NOISEX database with noise and channel distorted speech are presented.,
Original languageEnglish
Title of host publicationProc. 4th European Conference on Speech Communication and Technology
Publication statusPublished - 1995
EventEurospeech 1995 -
Duration: 1 Sep 1995 → …


ConferenceEurospeech 1995
Period1/09/95 → …

Cite this