INTERSPEECH 2006  ICSLP

In a previous work, MultiEnvironment Model based LInear Normalization, MEMLIN, was presented and it was proved to be effective to compensate environment mismatch. MEMLIN is an empirical feature vector normalization which models clean and noisy spaces by Gaussian Mixture Models (GMMs). In this algorithm, the probability of the clean model Gaussian, given the noisy model one and the noisy feature vector (crossprobability model) is a critical point. In the previous work the crossmodel probability was approximated as timeindependent. In this paper, a timedependent estimation of the crossprobability model based on GMM is proposed. Some experiments with SpeechDat Car database were carried out in order to study the performance of the proposed estimation in a real acoustic environment. MEMLIN with timeindependent crossprobability model reached 70.21% of mean improvement in Word Error Rate (WER), however, when timedependent crossprobability model based on GMM was applied, the mean improvement in WER went up to 78.47%.
Bibliographic reference. Buera, Luis / Lleida, Eduardo / NolazcoFlores, Juan A. / Miguel, Antonio / Ortega, Alfonso (2006): "Timedependent crossprobability model for multienvironment model based LInear normalization", In INTERSPEECH2006, paper 1271Wed1BuP.2.