- Master Thesis in Computer Science -
Visualizing sound for the deaf
This thesis describes the first steps in the development of a visual representation of auditive information for the deaf.
With this, deaf people can 'see' sounds, and even understand speech.
Although this thesis intends to describe only the first steps, small-scale experiments have already shown that the concept works, and people can recognise many features of sounds, including slowly spoken speech, after only a few minutes of training.
It's far less radical than a cochlea implant - and it's very easy to replace as technology gets better.
If you are not interested in the mathematical background, just skip chapter 2.|
If you lack the necessary mathematical background, you can just read the rest of the text and view the images. Especially the images will clear up a lot.
To get an overall idea of what is done here, view the AVI-files below. Put the two legenda-images on your screen, and with that try to interpret each image in the AVI-files.
DOWNLOAD THESIS AS A POSTSCRIPT FILE (zipped, 4.3 MB)
(PREFERRED) - Warning: Size after unzipping is about 160 MB
READ THESIS ONLINE (HTML version)
Symbols in the HTML version may be displayed incorrectly on some systems.
If you are capable of viewing PostScript files, download the PostScript version!
The HTML-file was automatically generated from a LaTeX-file by the program TTH, version 2.24.
DEMO .AVI FILES
VOICE SAYING "SHT!"
VOICE SAYING "OI UA"
VOICE SAYING "ONE MAN LIVING ON AN ISLAND"
INSTRUMENTAL MUSIC Not available online due to copyright issues
In order to understand these AVI films, take a look at the legenda:
PART 1 and PART 2.
DOWNLOAD THE SOFTWARE
C-sources (tested for DOS, Unix and Linux with GNU C-compiler).
Please send your comments to firstname.lastname@example.org