Working on speech recognition

Miscellaneous tools
Post Reply
Message
Author
stemsee

Working on speech recognition

#1 Post by stemsee »

I have managed to obtain directly from the German programmer source code for speech recognition/assistant software. When i have it working I will post something.

amigo
Posts: 2629
Joined: Mon 02 Apr 2007, 06:52

#2 Post by amigo »

Pant, pant... at least give us the name so we can be dreaming about it meanwhile...

stemsee

#3 Post by stemsee »

It is called simon ... I have had it sitting here for a few months. A dependency is kde, that is why i waited so long, I was going to build a kde only pup. EmSeev1 had kde, LH64 has kde.sfs so maybe I will use that to build it.

Code: Select all

Please compile Simon with the provided build.sh script.

There are the following dependencies.

Required build-time dependency:
1. KDE libraries and workspace (4.8 or newer)

Optional build-time dependencies:
1. KDE Pim Libs . Enables the optional command plugin "Akonadi" that allows
Simon to trigger commands at certain times and to use Simon dialogs as
calendar reminders.
2. LibSampleRate (http://www.mega-nerd.com/SRC/). If you build Simon with LibSampleRate support, it can
automatically resample audio in- and output for devices that don't support the
required configuration.
3. qaccessibilityclient (http://cmusphinx.sf.net://projects.kde.org/projects/playground/accessibility/libkdeaccessibilityclient). Enables the experimental ATSPI integration.
4. Sphinxbase, PocketSphinx, SphinxTrain (http://cmusphinx.sf.net). Together, they enable SPHINX support in Simon
(model compilation and recognition).
5. OpenCV (http://opencv.org). Enables the face detection and lip detection context plugins.

Optional runtime dependencies:
Julius (http://julius.sourceforge.jp/en_index.php) and the HTK
(http://htk.eng.cam.ac.uk) enable the use of HTK models when installed.


Please refer to the manual for more information.

User avatar
rufwoof
Posts: 3690
Joined: Mon 24 Feb 2014, 17:47

#4 Post by rufwoof »

There are some demo's (for Simon) on youtube if you search around. I watched one only a week ago or so. I like how it divides up the screen into numbered regions and you say a number and it then subdivides that region into numbered sections .... etc until you've zoomed in on the area/icon/whatever you had in mind and can verbally activate that icon/whatever. OK for limited use I guess, but I suspect that would be less acceptable if you were doing lots of clicking around.

User avatar
ETP
Posts: 1193
Joined: Tue 19 Oct 2010, 19:55
Location: UK

Re:Working on speech recognition

#5 Post by ETP »

Hi stemsee,

You may wish to check out LiSpeak which appears to be under development.
It uses as its base Palaver that I used in a butchered form in Obedient Pup.
https://github.com/BmanDesignsCanada/LiSpeak/
Regards ETP
[url=http://tinyurl.com/pxzq8o9][img]https://s17.postimg.cc/tl19y14y7/You_Tube_signature80px.png[/img][/url]
[url=http://tinyurl.com/kennels2/]Kennels[/url]

amigo
Posts: 2629
Joined: Mon 02 Apr 2007, 06:52

#6 Post by amigo »

Okay, nothing new. Simon has been around quite awhile, and as mentioned centers on kde so it is naturally heavy. Palaver looked good, although it is pretty heavy since it uses python. I think it has been replaced by a new project.

Simon makes a good case study because it is pretty functional and well thought-out. But, to me, only for ideas.

Post Reply