This course discusses the concept of speech processing and some of its applications. This includes; acoustic theory of speech production, discrete time models of speech production, time and frequency domain analysis techniques. It also discusses approaches in speech analysis, speech synthesis, speech coding and more applications on speech and audio signals.
Course Objectives:
The main purpose of this course is to provide the learners with solid fundamental tools and experience on the applied technology in the array of speech processing. It also gives the opportunity to deepen and widen the understanding of this topic through the study of speech characteristics and its applications in various fields like coding. The course is one of the interesting applications of signal processing.
Intended learning Outcomes and Competences:
At the end of this course students should be able to;
Analyze and design algorithms for extracting parameters from the speech signal.
Analyze and design algorithms for speech and audio coding.
Topic Two :Introduction to Speech Coding What, Why, Where & How
Introduction:
It is aimed hear to write the speech signal in a proper format suitable for transmission or storage. Being speech, this signal has special characteristic which enables us to remove redundancy. Speech coders are classified into several categories and Speech Coding Hierarchy is discussed. Mobile Speech coder is introduced as an example.
Third: THIS IS AN EXTRA TOPIC. It shows the channel coding concept of Mobile System.
Study the following animation to define the term "channel coding", describe the channel coding proces, distinguish the different speech bit classes and identify the significance of control channel signaling data . Source coding removes redundancy but channel coding increaes redundancy
Analogue signal is sampled then it becomes discrete in time but not digital. Quantization is used to discretize the amplitudes of theses samples. Quantization introduces some error and so SNR at the output of the quantizer is considered.
Second :Watch the following video about what is quantization
Third :See the following video to understand quantization
Fourth :Use the following animation which illustrates the quantization of an analog signal. The original signal could be voice, audio, music or any other analog signals that may be converted into digital wavefrm .
An important application of speech coding is used in mobile communication systems. It is required to compress the speech signal to meet the Mobile channel limitations and resources restrictions.
Objectives:
Lable the components of sppech coder used in mobile communications
Speech compression is an important application of speech processing. Redanduncy should be removed from the signal. Several compresion methods are presented.
Natural speech production mechanism may be modeled in several ways. Here, we will check see how to analyse the speech using certain windowing methods, then it is required to study some models of Speech production
Speech parameters are to be quantized and encoded properly. Hence, LPC parameters may not be suitable for all kinds of processing. Hence, other forms of such parametrs should be considered
Objectives:
Acquire the importance of other forms of coefficints.
Being able to transfer the cofficients int other forms like LAR, PARCOR,..