Public lecture

Samuli Siltanen works as a Professor of Industrial Mathematics at the University of Helsinki, Finland, and is a team leader in the Centre of Excellence in Inverse Problems Research granted by the Academy of Finland for the periods 2006-2011 and 2012-2017. He has also worked as a research and development scientist at medical imaging technology companies, including GE Healthcare, and is President of the Finnish Inverse Problems Society.

Tuesday 26th August 6-7pm.
Understanding Human Speech

Human speech is the most sophisticated and complex means of communication to have ever existed and plays an unparalleled role in today's society.

Its complexities have informed the study of phonetics and linguistics and inspired widespread research in neuroscience and engineering. Whether developing the latest voice recognition software for a smartphone or designing computers to aid people who have lost their voice through disease and illness, researchers are finding ways to map the precise mechanisms in the human vocal tract that create the almost infinite number of words and phrases we hear every day.

This public lecture, given by Professor Samuli Siltanen from the University of Helsinki, will describe how a technique called glottal inverse filtering (GIF) is being used to perfect synthetic speech.

Through a number of practical demonstrations, Professor Siltanen will explain how the human vocal folds and the mouth and lips combine to create the vowel sound and how GIF can be used to determine the exact mechanisms used by these body parts by simply measuring the signal from spoken sounds.

The extensive use of speech also means that the vocal folds can come under severe stress. Outside of the obvious sounds of a weak and croaking voice, this 'vocal loading', which is often experienced by actors, singers and teachers can be very difficult to measure objectively.

In this lecture, Professor Siltanen will also describe how an emerging technique, known as electrical impedance tomography (EIT), is being tested as a reliable method for measuring when the voice starts to become overused.

EIT involves tiny electrodes being placed on the surface of a tissue or structure to measure its electrical conductivity or permittivity. Algorithms can then be used to reconstruct these measurements and form an image of the condition of the vocal folds.