Design and application of acoustic icons

OData support
Supervisor:
Dr. Csapó Tamás Gábor
Department of Telecommunications and Media Informatics

In this paper I look into how computer-generated, text-to-speech synthesized acoustic icons can be used in real-life environment.

I explore the methodology, history and characteristics of the acoustic icons, with special emphasis on their development. I research one of the state-of-the-art computer generated icon types in more detail. I develop a set of customized icons on both computer generated male and female voice, than attempt to record their human uttered versions so I can evaluate on their mutual commutability. That is to mark the quality of the computer-based voice synthesis.

For the sake of the simulation of real-life circumstances I put these icon sets all into one custom designed working prototype system to demonstrate their usability in practice. I develop a weather reporter Android software dedicated for this purpose, which I take to a group of testers to evaluate.

My ultimate goal with all of this is to find out if these computer-generated sounds are similar to human sounds in the matter of their capacity to express feelings and thoughts. I seek to know if they can transport relevant information to the user enhancing or even replacing the visual experience. But most of all I want to determine if they can get the sympathy of the users before the human recorded sounds. If so, that means these sounds are applicable for real-life usage and can replace human voice recordings.

Downloads

Please sign in to download the files of this thesis.