Interspeech'2005 - Eurospeech
Many forms of time varying acoustic models have been investigated for speech recognition. However, there has been little success in applying these models to Large Vocabulary Continuous Speech Recognition (LVCSR). Recently, fMPE was introduced as a discriminative feature space estimation scheme for the HMM-based LVCSR. This method estimates a projection matrix from a high dimensional space (กซ 100,000) down to a standard feature space (typically 39). This projection is then added on to the original feature vector (e.g. MFCC or PLP) to yield a feature vector to train the final model. This paper considers fMPE as a time varying model for the mean vectors by applying the time varying feature offset to the Gaussian mean vectors. This approach naturally yields the update formulae for fMPE and motivates an alternative style of training systems. This concept is then extended to the temporal precision matrix modelling (pMPE). In pMPE, a temporally varying positive scale is applied to each element of the diagonal precision matrices. Experimental results are presented on a conversational telephone speech task.
Bibliographic reference. Sim, K. C. / Gales, M. J. F. (2005): "Temporally varying model parameters for large vocabulary continuous speech recognition", In INTERSPEECH-2005, 2137-2140.