ViaVoice and XVoice: Providing Voice Recognition

An overview of voice recognition software for Linux, its benefits and shortcomings and its cloudy future.
Issues and the Future

Some applications, mainly games such as TuxRacer, bypass X for key presses, leaving XVoice unable to control them. Mouse-heavy applications, such as The GIMP or Netscape, can be automated, but it's extremely tedious to try to control the mouse by voice. Fortunately, Mozilla 1.2a has "type ahead find", which, in conjunction with XVoice, lets you speak text-within-a-text-link to navigate web pages by voice.

Voice recognition in general works great for commands and fine for casual text. However, even small error rates can be quite annoying for some uses. Be advised that it can be exasperatingly difficult to program by voice. Another issue to be aware of is the possibility of straining your voice, much as it is possible to overuse your arms.

While XVoice and ViaVoice put a lot of power at your control, it is not quite possible to control entirely the Linux desktop by voice. This is disappointing to anyone needing hands free accessibility. Sadly, the weak link is IBM. At least with the version shipped by IBM, Dictation requires keystrokes for unavoidable dialog windows, for example, and Dictation is the only way to train ViaVoice. Of course, if you don't need any additional training, can automate all your applications and aren't concerned about security, you're in good shape.

IBM has released new versions of ViaVoice for Mac and Windows but not for Linux. Despite all the money they're spending on other areas of Linux, they don't even actively market ViaVoice, and their future support is unclear. In March they pulled without comment the ViaVoice Linux SDK, which XVoice needs to compile. With this cloud over the future, the XVoice developers are currently trying to find a viable open-source alternative instead of adding new features. A group of developers and users is out there wanting to make ViaVoice for Linux a success, but without even minor support from IBM the opportunity will be missed.

Rob Spearman is a Seattle software architect recovering from an overuse syndrome. This article was written using voice recognition on Linux.

email: rob@smeg.com

______________________

Comments

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

IBM ViaVoice

Anonymous's picture

I bought ViaVoice from IBM for about $40 in the Spring of 2002. It was the worst install I ever seen! if it wasn't for internet posts I would have never gotten it working. I doesn't come with the needed black down java. You need an older version because thats all its been tested with.

No manual, just a CDROM, headset and a folder sheet of paper called instructions. I had it running on SuSE 7.2, but after installing Jbuilder it broke and I never used it much anyway. Atleast you get a headset. The audio levels are very very low. I bought the headset for OS/2 many years ago and voice type dictation (pausing in betwen words) was more impressive on my old P90 compared to ViaVoice on my Athlon 1900mhz system! The joy of Java intrepretted code! When CPUs get faster maybe thell come out with a GWBASIC version! Anyway, this product is love / hate. The continious speech works nicely, but the error rates are high one error in about 10-20 works. 95% sounds good but thats one wrong word in every 20. Atleast the words are all spelled correctly.

The gui is sluggish and pretty basic. I expected more from IBM. I seen better 0.x open source projects!

--Ed March

http://www.poetworld.org/~emarch

Re: IBM ViaVoice Vs other products

Anonymous's picture

Voice recognition has greatly improved. I have ViaVoice
for os X and WinXP and think it is great but could be
much better.

How does the CMU or other offerings compare with
these two well known standards?

I appreciate knowing that the ViaVoice for Linux isn't
worth loading. Maybe that is why IBM fired all the
linux developers and took the nix offering off their
website.

Link to IBM purchase page

tjmather's picture

Here's the link to purchase the IBM Viavoice for Linux.

Also, there is a new xvoice-sphinx project.

Re: ViaVoice and XVoice: Providing Voice Recognition

Anonymous's picture

IBM took the Linux page off its web server, but if you want to buy it you can still order it over the phone at IBM Direct:

800-426-2255.

Also, if you are concerned about the future of Linux voice recognition, get involved.

Join a mailing list and keep up to date.

Harass IBM. Current law REQUIRES that businesses make reasonable accomodation for disabled employees, and VR software will soon be a mandatory offering of each business. It will not be possible for them to make inroads in the office desktop market without it.

sphinxTrain :(

Anonymous's picture

I am working on a project which is using sphinx II as speech recognition. But I am too stupid to understand the its training produre (although I have read its documentation over and over again) so if anyone who know how to train sphinx II using sphinxTrain please email me at vinhtran@ork.net

Please clarify me how the work go on step by step and how to set up things for it (the more detailed the better)

I know this will take a lot of your time but...PLEASE HELP ME. I WOULD HIGHLY APPRECIATE.

Thank in advance.

ViaVoice for Linux is still available!

Anonymous's picture

According to IBM, ViaVoice for Linux is available over the phone at 800-426-2255. For some reason they have removed it from their web store. If you're tempted to try out serious voice recognition on Linux, I suggest you purchase a copy and let IBM know that there is a market for the Linux version.

No thanks, I'm sticking with

Anonymous's picture

No thanks, I'm sticking with the idea of open source on linux, and padding IBMs pocket isn't appealing to me. Maybe sphynx will work out, or i'll just keep typing.

Re: ViaVoice and XVoice: Providing Voice Recognition

Anonymous's picture

What about CMU Sphinx?

nobody ever seems to realize that there IS an open-source continuous flow speech recognition engine out there.

actually there is two

CMU Sphinx-II, which deals with limited domain speech, and

CMU Sphinx-III, which deals with unlimited domain speech for dictation purposes.

google for them.

Re: ViaVoice and XVoice: Providing Voice Recognition

Anonymous's picture

In March they pulled without comment the ViaVoice Linux SDK, which XVoice needs to compile.

So, like, then what's the point of this article? "There's this cool technology but it's no longer available and old versions are no longer supported" is hardly a tutorial or "that it is possible to control the Linux desktop by voice, and it has been possible for some time" -- this is just Linux propaganda: The correct statement is that it was almost available under linux, but the vendor lost interest before the product was viable. For what its worth, numerous postings to the IBM ViaVoice list have pleaded for a contact address where we can purchase this mythical $40 Linux kit; if the author knows something the IBM engineers don't about where to get this, please post it -- that would be information worth publishing!

Re: ViaVoice and XVoice: Providing Voice Recognition

Anonymous's picture

If you download the XVoice RPM, you don't need the SDK. (You will need ViaVoice of course.) I have been using this set up for 8 months. With more users or more interest, IBM may have continued to support ViaVoice on Linux. It appears that the article came out just after they pulled ViaVoice for Linux. Bad timing.

Re: ViaVoice and XVoice: Providing Voice Recognition

Anonymous's picture

Not just propaganda. Check out cvoicecontrol:

It converts spoken words to commands to be executed. It's a little dated and not maintained anymore, but it's available and I've found it work nicely.

ViaVoice/XVoice isn't the only piece of software capable of controlling the computer by voice.

Re: ViaVoice and XVoice: Providing Voice Recognition

Anonymous's picture

Hi!

How easy is it to use the CVoiceControl with standard Soundblaster and pentium III 450 MHz CPU?Could it recognize the word "One" and execute the command "ls"?

IBM Does Support Linux

Anonymous's picture

IBM Linux Portal

But IBM is not a charity with unlimited funds able to afford to

keep every product going forever, even if no one buys it.

Clearly people are buying Win and Mac versions of ViaVoice.

I guess not enough people were willing to pay actual money

for ViaVoice for Linux.

Re: IBM Does Support Linux

Anonymous's picture

comon now... open source demagogy is not about "charity" it's about developement and that's why IBM should provide the SDK version at least. They ARE a megacorp but they risk getting worse...

Well done IBM! May bancrupcy hit your stands!

Re: ViaVoice and XVoice: Providing Voice Recognition

xtifr's picture

At a recent Linux tradeshow, IBM was demoing some products using ViaVoice, and they had a few copies of ViaVoice they were giving away in drawings. I asked if they had any copies for sale, and they said no. So I asked if they had any literature or sales brochures, and they said no. I asked if there was any way to get more information, and nobody in the booth knew. I walked away, bemused.

Re: ViaVoice and XVoice: Providing Voice Recognition

Anonymous's picture

IBM doen't support Linux anymore, they fire all the Linux staff, and don't sold or distribute any Linux Software.

Support Linux, don't buy IBM stuff

Re: ViaVoice and XVoice: Providing Voice Recognition

Anonymous's picture

It's true. And Microsoft hired all the ex-IBM Linux people to produce Office for Linux, which will be out next month. Also, Elvis lives in my dog Buddy's dog house! Stupid elvis! I kick him out, but he keep moving in!

ViaVoice for Linux no more

Anonymous's picture

I started to play with ViaVoice a few times. I sure don't mind paying for software, but I hate the thought of depending on proprietary stuff for anything important. It could go *poof* at any time. Looks like that time has come for ViaVoice for Linux.

I'd like to get some *simple* voice recognition for a limited (aviation) vocabulary.

Re: ViaVoice for Linux no more

Anonymous's picture

Try out CMU Sphinx...

http://fife.speech.cs.cmu.edu/sphinx/

You can create your own simple vocabulary pretty easy. With some of the ones I have made, it is so good at recognizing that it is scary...

Call up 1-877-268-7526 and give it a try.

Later,

Ryan

No ViaVoice!

Anonymous's picture

Good article, but from what I can see IBM no longer offers any version of ViaVoice for Linux. Unless they've carefully hidden it away on their site, that is.

ViaVoice

Anonymous's picture

So, where is possible to find the free ViaVoice rpm version for Linux? I try to find it on the net, but all the link are to the IBM's page.

google

Anonymous's picture