Exploring sound in real time using MATLAB: a new tool for research and education Seminar

When working with sound, it is very often helpful to get instantaneous feedback about sound properties. Looking at a spectrogram is an invaluable tool for teaching about speech, and this even more so when the student can see the live spectrogram of their own voice.

I have published a platform that allows to investigate a multitude of sound representations in real time in a visual form. Furthermore, the student can explore complex signal processing algorithms to modify the real time sound, for example using a cochlea implant simulation, a hearing aid or a hearing loss.

Finally, it is possible to measure many objective aspects of the sound, loudness, sharpness, annoyance, pitch, speech intelligibility and quality and many more.

The platform can be downloaded from the Mathworks website .

In the presentation I will give a run through demo showing some aspects. I will furthermore briefly explain how anyone can add their own real-time modules.

my research spans the bridge from biology to engineering. I am interested in speech and the brain. The aim of my research is to understand how we hear sound and to use this knowledge to create bio-inspired solutions to enhance speech communication.

• Machine learning techniques to enhance speech
• Bioinspired auditory features for machine learning applications
• Auditory modelling and perception
• Neurophysiology of the cochlear nucleus
• Speech perception in noise
• Hearing impairment and Audiology
• Signal processing to enhance speech intelligibility in noise
• Building a low-cost, self-programmable hearing aid
• Improving user satisfaction of hearing devices with smartphone control