The vast majority of HMM–based speech recognition systems use Gaussian mixture models as the state distribution model. The use of these distributions is motivated more by ease of training, decoding and the fact that a sufficient number of Gaussian components may be used to approximate any distribution, than some underlying aspect of the data being modelled. If distributions were selected that b...

We propose a sophisticated tying mechanism for modeling deletion transfonnations between dialects. We empirically show that the proposed tying mechanism reduces deletion errors by 33% when compared to a baseline system using a standard tying mechanism. Statistical tests show that the proposed and baseline models make statistically diOcrcnt errors, thus suggesting that they are complementary sys...

Context-dependent modeling is a widely used technique for better phone modeling in continuous speech recognition. While different types of contextdependent models have been used, triphones have been known as the most effective ones. In this paper, a Maximum a Posteriori (MAP) estimation approach has been used to estimate the parameters of the untied triphone model set used in data-driven cluste...

Tying of Hidden Markov Model states is an important issue for the use of triphones as modeling units in automatic speech recognition systems. This paper studies the application of a–priori rules for tying in combination with data driven methods. The baseline method features a combination of a–priori rules that reduce the theoretical number of units by an oder of magnitude and a simple back–off ...

Performance of any continuous speech recognition system is highly dependent on performance of the acoustic models. Generally, development of the robust spoken language technology relies on the availability of large amounts of data. Common way to cope with little data for training each state of Markov models is treebased state tying. This tying method applies contextual questions to tie states. ...