Almost There!!! Today is my presentation at GUADEC.

I will talk about speech recognition and accessibility.

The talk is: Using the Speech as a Tool of Accessibility [1]

[1] http://www.grancanariadesktopsummit.org/node/214

I’m glad to announce I was accepted to make a presentation at Guadec 2008. I will make the presentation together with Nickolay V. Shmyrev who was my mentor on Google Summer of Code last year.

More information about our presentation can be found at [1].

[1] http://live.gnome.org/GnomeVoiceControl/Guadec2008

I made a Benchmark of Sphinx2, Sphinx3, and PocketSphinx to analyze the memory usage, time to decode, and errors on recognition ( errors on sentences and on words ).
The tests were made on a AMD Athlon(TM) XP 2000+ 1670.608 MHz with 512MB of memory.

Sphinx2

Memory used: 4.7% of 512MB = 24,064MB

root@controle03# time perl scripts_pl/decode/slave.pl
MODULE: DECODE Decoding using models previously trained
Decoding 130 segments starting at 0 (part 1 of 1)
Using files: 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Finished
SENTENCE ERROR: 78.462% (102/130) WORD ERROR RATE: 37.904% (293/773)

real 0m49.394s
user 0m46.599s
sys 0m0.296s

Sphinx3

Memory used: 5.8% of 512MB = 29,696MB

root@controle03# time perl scripts_pl/decode/slave.pl
MODULE: DECODE Decoding using models previously trained
Decoding 130 segments starting at 0 (part 1 of 1)
0%
SENTENCE ERROR: 80.8% (105/130) WORD ERROR RATE: 34.7% (268/773)

real 5m47.224s
user 5m36.393s
sys 0m1.932s

PocketSphinx

Memory used: 0.6% of 512MB = 3,072MB

root@controle03# time perl scripts_pl/decode/slave.pl
MODULE: DECODE Decoding using models previously trained
Decoding 130 segments starting at 0 (part 1 of 1)
0%
SENTENCE ERROR: 78.5% (102/130) WORD ERROR RATE: 37.9% (292/773)

real 0m0.211s
user 0m0.156s
sys 0m0.052s

With this benchmark we can see that Sphinx3 has less errors on recognition of words ( WORD ERROR RATE ) than Sphinx3 and PocketSphinx, but the time of decode is much bigger compared with Sphinx2 and PocketSphinx. The memory usage of Sphinx3 is greater than Sphinx2 and PocketSphinx.

The PocketSphinx has less memory usage and is the faster decoder.

The Gnome-Voice-Control is been modified to use PocketSphinx. Soon we’ll have the release 0.3 using PocketSphinx.

Thanks,
Raphael Nunes

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.

Hi, I made a brief demonstration’s video to show the functionalities of the Gnome Voice Control First Release that is been developed on GSoC.

I hope that you enjoy.

I would like to thank luizirber more known as Gaucho, by the help on making the demonstration video.

The software can be found on:

http://www.comp.ufscar.br/~raphael18/gnome-voice-control-0.1.tar.gz

Instructions to instalation can be found on:

http://raphaelnunes.wordpress.com/2007/06/07/gnome-voice-control-01-release-gsoc/ 

Thanks,
Raphael Nunes

First I will write a brief introduction about me:

My name is Raphael Nunes, I’m student of Computer Engineering at Federal University of São Carlos (UFSCar). Currently I am working with SYS administration at SimuCAD (Department of Production Engineering – UFSCar) and I’m participating on Google Summer of Code 2007 working on a GNOME Project.

In this first post I would like to announce the first release of the software that I ‘m developing in the Google Summer of Code 2007 (GSoC). To whom don’t know GSoC enter in the page http://code.google.com/soc. The software is gnome-voice-control and is being developed to Gnome Foundation. The project application that I submitted to GSoC can be found at http://code.google.com/soc/gnome/appinfo.html?csaid=4F64D394968BB092.

The gnome-voice-control is an applet software developed to control the GNOME Desktop by voice.

The gnome-voice-control can be downloaded from http://www.comp.ufscar.br/~raphael18/gnome-voice-control-0.1.tar.gz

To install gnome-voice-control:

First download sphinx2 that is a speech decoder developed by Carnegie Mellon University.

wget http://cmusphinx.org/download/nightly/sphinx2.nightly.tar.gz

Extract

tar -xvvzf sphinx2.nightly.tar.gz

Enter in the sphinx folder

cd sphinx2

Install the sphinx with the commands as root

./configure –prefix=/usr

make

make install

Exit the sphinx2 folder

cd ..

Once sphinx2 installed, download gnome-voice-control

wget www.comp.ufscar.br/~raphael18/gnome-voice-control-0.1.tar.gz

Extract

tar -xvvzf gnome-voice-control-0.1.tar.gz

Enter in gnome-voice-control folder

cd gnome-voice-control-0.1

To install gnome-voice-control execute the following commands as root

./configure –prefix=/usr –with-sphinx2=/usr

make

make install

Now, the gnome-voice-control is installed.

Add the gnome-voice-control to gnome-panel

Click with the right button at the panel and choose add to panel, then will appear a window to choose applets. In the Micellaneous section has an applet with name VoiceControl, add it.

After all, you will see an applet with a text ‘idle’. To make the gnome-voice-control function, click with the right button on the applet and choose start control. When the applet show the text ‘ready’ you can start to speech.

To make the gnome-voice-control stop, click with the right button on the applet and choose stop control.

I hope that you like the software.

The actions implemented in this first release are:

RUN TERMINAL : Open gnome-terminal

RUN MAIL : Open Evolution

RUN BROWSER : Open Epiphany

MINIMIZE WINDOW

MAXIMIZE WINDOW

CLOSE WINDOW

NEXT WINDOW

Comments about the software are appreciated.

Thanks,

Raphael Nunes