SPEECH SIGNAL PROCESSING

Abstract:

The application presentation deals with bringing out the importance of basic concepts of Signals and Systems in the field of speech communication and it’s contribution to the technological advances.

Introduction

Speech processing typically involves a basic representation of a speech signal in a digital domain which requires limiting the band width of the signal, sampling it at a certain corresponding rate and storing each sample with an adequate resolution.

But our focus in the field of speech processing is in communication. Speech can be represented in terms of a signal carrying some message content or information.

Speech signals can be thought of signals in both continuous and discrete domain.

• Speech signals are produced by our vocal cords and are transmitted from speaker to listener due to the pressure waves that are propagated through air resulting in the form of continuous or analog signals.
• Speech when stored as digital information in computers are represented in the form of discrete signals i.e. using integers as the independent variable.
The basic ideas behind speech communication system are
1. Transmission
2. Storage and
3. Processing

Generally all disciplines of signal analysis involve the following steps :-
• Production of the signal from an information source like Speech delivery
• Signal representation in natural domains
• Signal representation in transform domains to get better understanding of the concepts which are not observable in natural domain. Speech can be represented as a waveform or in terms of bits for digital storage.
• Finally extraction and utilization of the information. Example :- In the field of speaker verification the identity of the speaker is the information.

In conclusion, some of the aims in the field of speech processing is efficient speech transmission, economical digital storage and improved communications.