Gnome-Voice-Control 0.2 Demonstration

29 06 2007

Hi, we made a new release of the gnome-voice-control, and we made a demonstration video to show the new features added.

We had some problems of recognition on some actions over the gedit menu, I think that my english isn’t so good😦

I hope that you enjoy this release🙂

I would like to thank Hugo Poloni by the help on making this screencast.

The release can be found on:

http://live.gnome.org/GnomeVoiceControl/Downloads

Information on instalation can be found here

Additionally to the information on instalation above, to make work the gnome-voice-control 0.2 you need to enable the Assistive technologies on your linux, and the system variable LANG need to be set as LANG=C.

To enable Assistive technologies , go to Preferences/Acessibility/Assistive technologies and check the box “Enable Assistive Technologies”. You need to logoff to the changes have effect.


Actions

Information

34 responses

29 06 2007
bastianazzo

as we can see, recognize so many different word is a tough work without a decent calibration…wouldn’t it be better to make an even short training before using the software? it could be just short enough to improve the personal settings. moreover, I can see a sphinxtrain on the website
other two questions: why are you using sphinx2 instead of one of the next releases (3 and 4)? couldn’t it improve the speed to recognize commands? at the present time, it seems to be a bit slow, isn’t it.
last one: do you have a roadmap or something like that?
anyway…nice job!!

29 06 2007
MyName

I loved the “cut”-part😀
(you pronounce “cut” the same way we Dutch people do to say the Dutch word for vagina😉

Nice to see this software in action though!

29 06 2007
Control Gnome through voice « Black Celebration

[…] Gnome through voice Raphael Nunes ha pubblicato un video dimostrativo di gnome-voice-control, applet per il controllo vocale del Desktop Environment giunta alla versione […]

29 06 2007
encrypted

Wow, nice wallpaper, can you give link or name?

29 06 2007
BookeldOr

Great job, even if the recognition engine seems a bit accent sensitive😉 it does not work too bad. Isn’t it possile to add a little learning phase on the first run ?
Will you add speech recognition in editable widgets in the future ?

Also, what does it do when there are multiple actions possible with the same label ? (windows vista displays overlay labels with numbers to choose the one you want to activate, it may be the solution).

29 06 2007
Jon

nice work!

29 06 2007
Блог » Blog Archive » Гласовна контрола за gnome

[…] може да се управува со глас со помош на gnome-voice-control и за разлика од vista – […]

29 06 2007
Tom

Wow, looking very impressive so far. I’m looking forward to the day i dont need my hands to control my computer anymore! keep up the good work.

29 06 2007
maurolinux

i’ve installated this but when i add gnome-voice control to gnome panel i add the 0.1.0 version… why?

29 06 2007
nsh

> why are you using sphinx2 instead of one of the next releases (3 and 4)?

They actually are not different releases but a completely different recognizers. We use
sphinx2 now because it’s the fastest version.

> i’ve installated this but when i add gnome-voice control to gnome panel i add the 0.1.0 version… why?

Probably you should restart bonobo-activation-server or you have older version installed too. The best way is to check voice_control_applet binary and kill previous process if it running.

29 06 2007
nsh

> wouldn’t it be better to make an even short training before using the software?

It can be trained during process actually but this feature isn’t supported now

> last one: do you have a roadmap or something like that?

text edition commands, dictation, multilingual models. Please suggest your entry🙂

29 06 2007
finferflu

This is very cool, thanks for developing it🙂

29 06 2007
Top Posts « WordPress.com

[…] Gnome-Voice-Control 0.2 Demonstration Hi, we made a new release of the gnome-voice-control, and we made a demonstration video to show the new features […] […]

30 06 2007
psymon101

Wonderful work, holds great potential for the disabled community.

30 06 2007
bastianazzo

are you going to think about a “magic word” that makes the voice recognition start?
let’s say: the program is always listening, but it reacts only after a particular word. that one could be chosen to be a weird one, not so frequent in common language and very well understandable by the program.
you could think to let the user pronounce this word just before each command (sounds a little bit annoying) or before a series of command, or that it could activate the recognition for a limited time…
how difficult should it be to implement the training? i think it’s a key feature to make everything better

30 06 2007
tia1982

Great software!
Can you publish a change log for this new version?

30 06 2007
dor

Great software – finally a decent voice recognition solution for linux.
I did have some problems, though – after messing arount a lot with compiling the program, I’ve managed to run it – but I couldn’t use the gEdit commands as you showed on the video.

Can someone help me?

30 06 2007
nsh

> are you going to think about a “magic word” that makes the voice recognition start?
let’s say: the program is always listening, but it reacts only after a particular word. that one could be chosen to be a weird one, not so frequent in common language and very well understandable by the program.

Nice and sensible request. Although in current speech recognition technology there are better solutions like duplicated generic context-independent model and so on. I think we’d better move in that direction

>how difficult should it be to implement the training? i think it’s a key feature to make everything better

We’ll look into it, thanks.

> I’ve managed to run it – but I couldn’t use the gEdit commands as you showed on the video.

please make sure you are running C locale and you have accessibility technologies enabled.

1 07 2007
dor

Yes, the accessibility technologies are enabled.
How can I run the C locale? Is “export LANG=C” good enough?

Thanks🙂

1 07 2007
nsh

You can choose language in gdm or add export LANG=C in .bash_profile

2 07 2007
dor

Thanks, now it works🙂
Too slow, though…

2 07 2007
tia1982

I’ve seen this comment:

“Could you please show how to add more commands and change the existent ones so I could have firefox instead of ephiphany (for example) and could also try to make an italian localization?”

I think this will be a very interesting thing to do, so anyone can personalize any command..

2 07 2007
bastianazzo

thanks for the answers to my questions…
which is the C-file we should look at for the development? by following it from the very first time would allow us to better understand the program. in my understanding, it shouldn’t be that complex already, because most of the job is actually taken by the sphinx2 engine, am i wrong?
what do you mean by: duplicated generic context-independent model?
can you post a link?
have a good week

2 07 2007
Alphalpha’s Voice Control « Vai Na Lousa Chefe!!!!

[…] Gnome-Voice-Control-0.2 aqui. […]

4 07 2007
nsh

> which is the C-file we should look at for the development? by following it from the very first time would allow us to better understand the program. in my understanding, it shouldn’t be that complex already, because most of the job is actually taken by the sphinx2 engine, am i wrong?

Sorry, I don’t quite understand, everything is in tarball, you can look there if you are interested. Program is not very complex now indeed.

> what do you mean by: duplicated generic context-independent model?

Something like that have sense, but it’s not so easy to implement without free voice data:

https://sourceforge.net/forum/message.php?msg_id=4381329

7 07 2007
bastianazzo

thank you, this is much clearer now.
so, the goal is basically to make a second engine that is going to recognize all the words not belonging to the set of commands, and associate them with a sort of “do nothing”, instead of using the mostly predictable, but still totally wrong word in the vocabulary.
is it possible to make the vocabulary “context sensitive”? i mean, is it possible to a-priori exclude from the set of recognizable words, the one that are not belonging to the current possible set of commands?
i hope i was clear enough, sorry if i keep doing questions, but this project is really interesting

8 07 2007
nsh

> so, the goal is basically to make a second engine that is going to recognize all the words not belonging to the set of commands, and associate them with a sort of “do nothing”, instead of using the mostly predictable, but still totally wrong word in the vocabulary.

Exactly, but it’s mostly not investigated area currently. At least there are no samples for that neither in sphinx nor in other decoders.

> is it possible to make the vocabulary “context sensitive”? i mean, is it possible to a-priori exclude from the set of recognizable words, the one that are not belonging to the current possible set of commands?

Our language model (FSG) is already context-sensitive. Sadly it’s not possible to change it on the fly (it’s possible in sphinx4 though) but between utterances we change FSG according to current window. Of course it’s also an interesting domain of investigations.

>i hope i was clear enough, sorry if i keep doing questions, but this project is really
interesting

Please do. Probably it’s better to mail me and Raphael instead since it’s hard to track blog comments.

24 07 2007
bastianazzo

hi again!
i’m sorry, but I didn’t get your mail on Gnome Voice Control project page (I’m referring to nsh, since you’re the one who is always replying😉 ).
i just wanted to point out this post I went by:
http://smart-machines.blogspot.com/2007/06/free-voice-recognition-software-that.html
maybe you caught it too (it was via linuxtoday.com). in a comment, they refer to OpenAir:
http://mindmakers.org/openair/airPage.jsp
as a simple way to interface a C++/C# program to sphinx…
maybe you already know all about those things…

bye!

5 03 2008
Vincent de Vries

I agree with MyName.

Kut (announced as you said “cut”) is a rude dutch word for Vagina.
I really can’t wait to have this app installed. It sounds like lots of fun at the moment, and can be very usefull in certain circumstances.
Most of all: It gives me a reason (read: excuse) to say Kut in public places.

Keep up the good work!

7 04 2008
opax

Hello

Bravo!! It’s great!
I ll try to customize it for me because i am french…
is there a place where we can find how to build the FSG files?

Thanks! and keep working!

3 05 2008
hp

Wow! I have been waiting for that so long! Thanks and keep going.. best regards from germany..

10 06 2008
ngun

Wonderful!
I think you should also see if it can control a web browser as it is the most used application for most people.
open
bookmarks
choose bookmark
scroll up/down
previous/next to move to those pages

would do for a start.

Hey, you guys deserve some donations and support!

9 08 2008
Juan

Hi!

I just installed gnome-voice-control on ubuntu Hardy (64bits) from Synapthic and qhen I run the applet I got an error message. I solve it folowing this instructions:

http://ph.ubuntuforums.com/showthread.php?t=839528

But, when it was finally running, it just keep on the calibration state, no more. How I can fix it? do I hv to do something else in order to get it ready?

22 12 2008
cncgeorge

I think the program is slick, I would like to find the available commands. Can someone tell where I can find them?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s




%d bloggers like this: