[Sugar-devel] TurtleArt Speech Recognition

[Rodrigo Parra]

>
> It seems to be a great project!!
> We have uploaded the code of the turtleblocks voice recognition plugin for
> Butiá and the documentation in
> http://www.fing.edu.uy/inco/proyectos/butia/mediawiki/index.php/Paleta_para_primitivas_de_reconocimiento_de_voz_en_tortuga
> Maybe can help for start thinking :)


Thanks Andrés, I've already started tinkering with your code. I am trying
to adapt it to use our central recognition process.
I got it to recognize commands (verified by logging results in the
final_result method), however, when it gets to the actual block loading I
get an error like this:

1404449506.201795 WARNING root: empezar
1404449506.202377 ERROR dbus.connection: Exception in handler for D-Bus
signal:
Traceback (most recent call last):
  File "/usr/lib/python2.7/site-packages/dbus/connection.py", line 230, in
maybe_handle_message
    self._handler(*args, **kwargs)
  File "/home/rparra/workspace/sugarlistens/sugarlistens/helper.py", line
41, in __final_result
    self.notify_listeners(text)
  File "/home/rparra/workspace/sugarlistens/sugarlistens/listened.py", line
45, in notify_listeners
    listener(text)
  File "/home/rparra/workspace/sugar-build/build/out/install/share/
sugar/activities/TurtleBlocks.activity/plugins/butia_listens/escucharButia.py",
line 85, in final_result
    self.tw.prim_load_block(orden)
  File "/home/rparra/workspace/sugar-build/build/out/install/share/
sugar/activities/TurtleBlocks.activity/TurtleArt/tawindow.py", line 4918,
in prim_load_block
    dy = int(self._find_block(name, pos[0], pos[1]))
  File "/home/rparra/workspace/sugar-build/build/out/install/share/
sugar/activities/TurtleBlocks.activity/TurtleArt/tawindow.py", line 4974,
in _find_block
    raise logoerror("#syntaxerror")
logoerror: #syntaxerror

I simply installed your plugin by copying it to TurtleArt's plugins folder.
Any ideas on what the cause may be? Do I need to install something else to
be able to actually insert one of your custom blocks?



> Yo intenté varias veces hacer reconocimiento de voz y nunca logré que me
> devolviera las ordenes correctamente.
> En particular mi idea era hacer reconocimiento de voz sobre sugar o sobre
> el navegador para que el usuario pudiera abrir aplicaciones, apagar la
> maquina o realizar búsquedas en el navegador.


 Ojalá tengas mejor suerte con Sugarlistens, Flavio =)

Aunque todavía estoy trabajando, con vocabularios reducidos se logran
resultados bastante buenos en mi opinión. Por ejemplo, ya conseguí abrir
algunas aplicaciones con la voz desde la vista de favoritos, e incluso
jugar a la Actividad Maze con comandos de voz.

Con respecto a la búsqueda por voz, en este momento es bastante más difícil
de implementar por la cantidad de opciones posibles para el reconocimiento.

Por si te interesa, te dejo algunos enlaces del proyecto:
https://github.com/rparrapy/sugarlistens
*https://github.com/rparrapy/sugar/tree/speech-recognition
<https://github.com/rparrapy/sugar/tree/speech-recognition>*

*https://github.com/rparrapy/maze/tree/speech-recognition
<https://github.com/rparrapy/maze/tree/speech-recognition>*

Cheers/Saludos,

Rodrigo







2014-07-02 23:33 GMT-04:00 Flavio Danesse <fdanesse at gmail.com>:

> Yo intenté varias veces hacer reconocimiento de voz y nunca logré que me
> devolviera las ordenes correctamente.
> En particular mi idea era hacer reconocimiento de voz sobre sugar o sobre
> el navegador para que el usuario pudiera abrir aplicaciones, apagar la
> maquina o realizar búsquedas en el navegador.
>
>
> 2014-07-01 10:18 GMT-03:00 Andres Aguirre <aguirrea at gmail.com>:
>
> It seems to be a great project!!
>> We have uploaded the code of the turtleblocks voice recognition plugin
>> for Butiá and the documentation in
>> http://www.fing.edu.uy/inco/proyectos/butia/mediawiki/index.php/Paleta_para_primitivas_de_reconocimiento_de_voz_en_tortuga
>> Maybe can help for start thinking :)
>>
>> Andrés
>>
>>
>> On Mon, Jun 30, 2014 at 8:39 AM, Rodrigo Parra <rodpar07 at gmail.com>
>> wrote:
>>
>>> Hello everyone,
>>>
>>> As a part of the GSoC project I am currently working on, Sugar Listens,
>>> I am planning to add speech recognition support to TurtleArt.
>>>
>>> The details of the implementation are still somehow blurry so, as
>>> discussed with Walter, I would like this thread to serve as a pool of ideas
>>> and debate around the subject.
>>>
>>> I've already been pointed out towards previous work developed at FING
>>> guys, who did an awesome job.
>>>
>>> The ideas mentioned during last week's meeting were:
>>>
>>>    - We need to define a proper grammar for TurtleArt. Every single
>>>    command counts, so debate is open.
>>>    - A speech related block, that would wait for a voice command for a
>>>    certain period of time during execution time. The behavior of such block
>>>    should be dynamic, meaning that actual actions triggered by it would depend
>>>    on the voice command.
>>>    - An activity to help users build a language model (grammar) would
>>>    be interesting too.
>>>
>>>
>>> Cheers,
>>>
>>> Rodrigo
>>>
>>>
>>> _______________________________________________
>>> Sugar-devel mailing list
>>> Sugar-devel at lists.sugarlabs.org
>>> http://lists.sugarlabs.org/listinfo/sugar-devel
>>>
>>>
>>
>> _______________________________________________
>> Sugar-devel mailing list
>> Sugar-devel at lists.sugarlabs.org
>> http://lists.sugarlabs.org/listinfo/sugar-devel
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sugarlabs.org/archive/sugar-devel/attachments/20140704/e0463b5b/attachment.html>