Matlab uses inbuilt sound card of the computer in order to record voice or sound through microphone connected to the system. 1: Block Diagram of a typical Audio/Speech Signal Processing
PLAY SOUND IN MATLAB HOW TO
Hence if we want to demonstrate such system using Matlab then we should also know about how to record sound using Matlab?įig. While working is real time speech processing needs to be done with live recorded sound. Once our system is trained to the specific user speech then we can apply it in real time.
In the above mentioned example first we need to train our system for specific user speech and that is done by extracting the features from pre-recorded voice samples and then train our system based on those features. You can think of this as words of a speaker are acting as the password for the valuable information. In speaker recognition problem our aim is to identify specific speaker and then provide authenticity of a task to specific speaker only.
Let’s take an example of speaker recognition. Prior method is used to do offline processing of speech and sound data in order to create algorithms for doing certain task, while later method is used to verify those algorithms and to perform tasks in real time. One is by reading pre-recorded sound files or data on computer and second is by recording sound in real time and then use that data in Matlab. As discussed in previous lesson and as shown in figure 1, there are two ways to process sound or speech information using Matlab.
0 kommentar(er)
