ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

The State of the Art of Automatic Speech Recognition: An Overview?

Journal: International Journal of Computer Science and Mobile Computing - IJCSMC (Vol.4, No. 2)

Publication Date:

Authors : ; ;

Page : 359-368

Keywords : Automatic Speech Recognition (ASR); Feature Extraction; Pattern Matching; Reference Pattern; Hidden Markov Model(HMM); Dynamic Time Warping(DTW); Multilayer Perceptron (MLP);

Source : Downloadexternal Find it from : Google Scholarexternal

Abstract

The objective of this paper is to present an overview of the techniques used in speech recognition systems. In this paper, we discuss the types of speech recognition showcasing the development in the field to help provide a technological perspective of the progress made in the field. Further, it highlights the fundamental principles and methods of Speech Recognition to understand the basic design required to build the technology. In addition, this paper discusses the various approaches to ASR and the classification techniques of the Speech Recognition System- HMM, DTW, MLP, along with their advantages and disadvantages. After decades of research, the efficiency of an ASR system and its accuracy remains the most crucial challenge. This paper attempts to review the basic technology of Speech Recognition, based on which we can build the most advanced systems overcoming the challenges we face currently.

Last modified: 2015-03-03 22:12:36