ABSTRACT
The task of speech recognition is to convert speech into a sequence of words by a computer program. As the most natural communication modality for humans, the ultimate dream of speech recognition is to enable people to communicatemore naturally and effectively.While the long-termobjective requires deep integration with many NLP components discussed in this book, there are many emerging applications that can be readily deployed with the core speech-recognition module we review in this chapter. Some of these typical applications include voice dialing, call routing, data entry and dictation, command and control, and computer-aided language learning. Most of these modern systems are typically based on statistic models such as hidden Markov models (HMMs). One reason why HMMs are popular is that their parameters can be estimated automatically from a large amount of data, and they are simple and computationally feasible.
