Speech Recognition: Theory, Applications, and Challenges

Speech is the most basic, common and efficient form of communication method for people to interact with each other. People are comfortable with speech therefore persons would also like to interact with computers via speech, rather than using primitive interfaces such as keyboards and pointing devices. This can be accomplished by developing an Automatic Speech Recognition (ASR) system which allows a computer to identify the words that a person speaks into a microphone or telephone and convert it into written text. As a result it has the potential of being an important mode of interaction between human and computers. Although any task that involves interfacing with a computer can potentially use ASR. The ASR system would support many valuable applications like dictation, command and control, embedded applications, telephone directory assistance, spoken database querying, medical applications, office dictation devices, and automatic voice translation into foreign languages.