An overview of speech recognition system based on the support vector machines

In the current scenario, speech recognition systems have become one of the premier applications for machine learning and pattern recognition technology. The speech recognition uses neural nets, hidden Markov models, Bayesian networks, DTW and other tools for recognizing the particular speech. The classification modules are playing a very important role in most of the modern speech recognition systems. The performance of the speech recognition systems is based on the classification techniques used for training the systems. Broadly, two methods are considered for developing the classification modules using either statistical methods or discriminative methods. To overcome from the occurred weaknesses, a new machine learning technique called support vector machine is introduced. This paper provides an overview of speech recognition system using support vector machines. Support vector machines (SVM) is a new approach to pattern classification that automatically control generalization, which gives good generalization and has been applied to various tasks, and parameterization as part of the overall optimization process. Specifically, SVM will be used to classify speech patterns.