Continuous Speech Recognition Technologies—A Review

Abstract

Speech recognition is the most emerging field of research, as speech is the natural way of communication. This paper presents the different technologies used for continuous speech recognition. The structure of speech recognition system with different stages is described. Different feature extraction techniques for developing speech recognition system have been studied with merits and demerits. Due to the vital role of language modeling in speech recognition, various aspects of language modeling in speech recognition were presented. Widely used classification techniques for developing speech recognition system were discussed. Importance of speech corpus during the speech recognition process was described. Speech recognition tools for analysis and development purpose were explored. Parameters of speech recognition system testing were discussed. Finally, a comparative study was listed for different technological aspects of speech recognition.