Slurred speech recognition system Files

This is the slurred speech recognition package. It consists of s2t which is
a speaker-dependent continuous-speech recognition program under development,
and soundView which is a speech waveform analysis program.

To start the speech recognition program s2t, execute

	java -Xmx2000m -jar s2t.jar

the -Xmx2000m option is to increase the maximum memory size of the
Macintosh JVM but is non-standard. A different flag may be used depending on
the host JVM and on how much memory is available on the host machine.

The voice recording can be stopped by simply clicking on the upper right
rectangle.

The source code for these programs is located in the file slurredXXX-src.jar.
Unpack the file with the command

	jar -xvf slurredXXX-src.jar

The code is compiled using the ant program http://ant.apache.org/.
Specifically, in the top level directory execute

	ant

To start the speech waveform analysis program soundView, go into the
directory slurred/tools and execute

	java -jar soundview.jar

There is a file in that directory with some instructions on how to use the
program.