Sorry, no experience.  But I’ve had the idea that voice compression to its 
basic elements (tone, pitch, energy) would make it easier for a recognition 
algorithm to operate.  There would be more core information and less useless 
artifact.  Seems like feeding a codec2 stream to an HTM (neural net)[1] would 
be one fun experiment.

[1] https://en.wikipedia.org/wiki/Hierarchical_temporal_memory 
<https://en.wikipedia.org/wiki/Hierarchical_temporal_memory>

!!Dean

> On Apr 12, 2018, at 1:54 AM, glen english <g...@cortexrf.com.au> wrote:
> 
> Anyone had any experience using codec 2 as a compressed storage method for 
> voice recognition.
> 
> Essentially, store the phrases in your FLASH as CODEC2-3600 frames.
> 
> do a search for a match.... re assemble the phonemes etc ...
> 
> glen VK1XX
> 
> 
> 
> 
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> Freetel-codec2 mailing list
> Freetel-codec2@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/freetel-codec2

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Freetel-codec2 mailing list
Freetel-codec2@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freetel-codec2

Reply via email to