Hi, Sorry for multiple postings, but I wasn't quite sure which list to direct this question to. I was wondering if anyone knows SuSE's current intentions towards Speech Recognition Software? I know that most of the distributions used to package ViaVoice, but this has now become defunct.I am interested to know if SuSE or KDE are going to try to work with any other recognition toolkit for linux in the near future. If SuSE or any of the large linux groups have chosen a speech recongition toolkit, then it tgives the rest of us in research a large indiction of which software to expect to become readily available. In some recent reading, I see CMU Sphinx as being one promising toolkit, but it is limited in a curent lack of multi-lingual support. In particular there is no readily available German Acoustic model, and a sizeable amount of effort would be required to put one together by trained linguists. I would be very grateful for any information or comments you might have on this topic. yours robert ross -- ==================================================================== Robert Ross -------------------------------------------------------------------- Mail FB3, Mathematik-Informatik Deptartment of Mathematics Universität Bremen & Computer Science Postfach 330 440 University of Bremen 28334 Bremen P.O. Box 330 440 Deutschland 28334 Bremen Germany Phone +49 421 218 7129 Fax +49 421 218-3054 Web http://www.informatik.uni-bremen.de/~robertr Email robertr@informatik.uni-bremen.de ====================================================================
On Tue, Jul 08, 2003 at 12:01:46PM +0200 or thereabouts, Robert Ross wrote:
I was wondering if anyone knows SuSE's current intentions towards Speech Recognition Software?
I don't know about SuSE's intentions, but I do have some thoughts / comments regarding this, as I have been involved in SR for over 10 years.
I know that most of the distributions used to package ViaVoice, but this has now become defunct.I am interested to know if SuSE or KDE are going to try to work with any other recognition toolkit for linux in the near future.
I believe the only OS that ViaVoice came bundled with was Mandrake, back in their 7.x series. ViaVoice for Linux is not "active" on IBM's website, but it is still available, and works with just about any distro. It works quite well, and I have successfully ported their other law and medical vocabularies from Windows, to run in Linux. I believe you are correct, that they did not come out with a German Acoustical model, but I believe that the Mandrake version did. Perhaps you can locate an older Mandrake version, and just install ViaVoice into SuSE. Really the only requirement for installation is that a certain version of Blackdown's Java is needed in order to properly run the "dictation pad." I had no trouble installing VV over several distros and versions. I have also been told, by an IBM programmer, who works with VV, that one can compile the source code for use in Linux, although I have never done this. Another possibility, and new finding is that if you have/use the newest v4.0 of VMWare, their default soundcard is now Soundblaster, and I run voice recognition of ViaVoice and Dragon Naturally Speaking in W2K in VMW. This was never possible before, and it works rather well, getting 99+% accuracy, which is the same as I achieved with VV for Linux.
If SuSE or any of the large linux groups have chosen a speech recongition toolkit, then it tgives the rest of us in research a large indiction of which software to expect to become readily available.
Point well taken.
In some recent reading, I see CMU Sphinx as being one promising toolkit, but it is limited in a curent lack of multi-lingual support. In particular there is no readily available German Acoustic model, and a sizeable amount of effort would be required to put one together by trained linguists.
Sphinx is also the one I have heard of also, and yes, it would take considerable effort and I do say, money to put together the acoustical and vocab models needed. -- Gary
On Tuesday 08 July 2003 15.31, gary wrote:
On Tue, Jul 08, 2003 at 12:01:46PM +0200 or thereabouts, Robert Ross wrote:
I was wondering if anyone knows SuSE's current intentions towards Speech Recognition Software?
Where can i find the ViaVoice packages?? I have looked for it in all kinds of places, but always failed finding it. I really would like to start with VV and speach recognicion. -- /Rikard ------------------------------------------------------------------------------------ Rikard Johnels email : rikjoh@norweb.se Web : http://www.rikjoh.com Mob : +46 70 464 99 39 ------------------------ Public PGP fingerprint ---------------------------- < 15 28 DF 78 67 98 B2 16 1F D3 FD C5 59 D4 B6 78 46 1C EE 56 >
On Sun, Jul 13, 2003 at 12:19:17AM +0200 or thereabouts, Rikard Johnels wrote:
On Tuesday 08 July 2003 15.31, gary wrote:
On Tue, Jul 08, 2003 at 12:01:46PM +0200 or thereabouts, Robert Ross wrote:
I was wondering if anyone knows SuSE's current intentions towards Speech Recognition Software?
Where can i find the ViaVoice packages?? I have looked for it in all kinds of places, but always failed finding it. I really would like to start with VV and speach recognicion.
If you can all into the US to an 800 toll free number somehow, Call 1-800-426-2255 x1968 Ask for Michelle Greenwood. give her the part # below. Below is taken from an email from someone who works at IBM, a few months back.. The ViaVoice for Linux (p/n 11K8437) can be purchased only through IBM Direct at 800-426-2255. This product is no longer available for sale through Shop IBM on the web. -- Gary
El Domingo, 13 de Julio de 2003 00:19, Rikard Johnels escribió:
Where can i find the ViaVoice packages?? I have looked for it in all kinds of places, but always failed finding it.
ViaVoice engine: https://www6.software.ibm.com/dl/viavoice/vvlinux-r?S_PKG=dl ViaVoice GUI: http://xvoice.sourceforge.net/
participants (5)
-
gary
-
gary
-
Rikard Johnels
-
Robert Ross
-
Sepio