Speech Signal Processing

We live in a noisy world. As a result, speech-user interface devices and assistive listening devices or hearing aids would be subjected to noise and fail to perform. An engineering solution to noisy speech signals is via speech signal processing, which involves the mathematical manipulation of speech signals. One example is speech enhancement, where noisy speech signal can be processed in such a way that the noise component is reduced, leading to an enhanced speech.

This workshop will give an introductory view to the world of speech processing with a particular focus on speech enhancement. The workshop will be divided into three parts, with the first session on the general introduction to speech processing methods. The second part of the morning session will be a case study on a successful speech processing spin-off company, Sensear Pty Ltd*. By using Sensear as a case study, this workshop will look at how a simple idea in a university environment can be brought to a commercial success. Audio demonstrations will also be presented where appropriate.

The final part (afternoon session) of the workshop will focus on the engineering aspects of audio signal processing by using a numerical software (MATLAB©) to construct a simple speech enhancement algorithm. Real life data in various noisy environments will be presented to the participants to experience a hands-on speech enhancing process. The algorithms can also be modified freely and rewritten as a mobile application.

*Sensear Pty Ltd is a multinational company which specializes in hearing protection, with its headquarters in Perth, WA Australia and global marketing headquarters in Larkspur CA, USA.

Who Should Attend

Academicians, researchers, industrial practitioners and students who wish to learn about optimization techniques and in real world applications.