Front-End Detection Technique of Speech Signal. An Experimental Study with reference to an Indian Language

      Аннотация: Language is considered as one of the most important medium of communication in a human society. Language in this sense is referring the human language, which is produce by the human speech tract in a normal condition. When we utter a word or a sentence in any language, linguistically we don’t speak the word or sentence, but we utter the phonemes in a sequence. So, to study a language for the use of modern communication technology, we have to study the characteristics of phonemes of that language The aim of the present book is to explain different types of feature extraction techniques of speech signal. Speech recognition consists of two stages. The first stage is known as front-end processing or feature extraction and the 2nd stage is the pattern matching process. The aim of this 1st stage of speech recognition is to use a priori knowledge to transform an input in the signal space to an output in a feature space to achieve some desired criteria. The feature extractor block designed in speech recognition aims towards reducing the complexity of the problem before the next stage start to work with the data.

Ключевые слова: Speech Signal Processing, Pitch, Formant Frequency, LPCC, MFCC, Speech recognition