Hello forks,
We are developing a natural language understanding system
which is using STk. This system has a virtual world, in which
virtual robots can act on user's instructions through the voice.
The user can see the resultant animation of the robots.
This system is build up of three modules, which are a speech recognition
module (with IBM ViaVoice), an animation generation module (with CMU Alice),
and a language analysis module (with STk). At the beginning, this system
supports only Japanese sentences though, now we implemented the English
version.
Currently actions which the robots can perform are quite limited (and
English grammar versatility is still in infancy). But we decided to
distribute our system on the web. See our web site:
http://tanaka-www.cs.titech.ac.jp/pub/kairai/
This distribution package runs on Windows 98 (as far as I know). This
package includes a STk binary and our system source written in Scheme.
The speech recognition module and the animation generation module are
not included. The system will work without these modules though,
you can obtain them freely (see the attached documents).
I am happy for hearing any comments.
Thank you
SHINYAMA Yusuke
--
Dept. of Computer Science, Tokyo Inst. of Technology,
2-12-1, Ookayma, Meguro-ku, Tokyo, 152-8552, JAPAN
E-MAIL: euske_at_cl.cs.titech.ac.jp
FAX: +81 3-5734-2915
Received on Sun Nov 26 2000 - 11:23:21 CET