Regarding google summer of code 2015

[Nico Kadel-Garcia]

On Tue, Feb 17, 2015 at 5:49 AM, Abhilash Mhaisne
<abhilashmhaisne at gmail.com> wrote:
> Sir / Madam,
>                    I am an engineering student, and gsoc aspirant. I want to
> develop a easy-to-use voice recognition system, which would be capable of
> judging even minor changes in accents, for fedora systems.

And I'd like a pony, with pretty wings that can fly to Jupiter and
makes cookies.

More seriously, voice recognition is one of the great challenges of
the last 50 years of speech analysis, approached by large and small
companies and research labs around the world. "Judging even minor
changes in accents" is a nightmare in real work: every commercially
available computer microphone system for the last few decades has made
the same basic mistake for collecting speech. They completely screw up
"plosives", sounds like "b" and "p" that have a lot of high frequency
information that gets completely munged by the digitization and
undersampling. And do not get me *started* on the ongoing fallacy that
"if we just collect more speech samples, we'll somehow be able to
analyze them". I'm aware of companies with software patents spending
many millions of investment money in the work, and they're still
getting their ass handed to them in the marketplace by smaller, local
apps that handle basic speech locally and *don't try to get fancy*.

And in case it's not clear: I did artificial hearing research for a
dozen years, designing electronics for cochlear implants. Every modern
microphone=>speech analysis system screws up the plosives, for a stack
of reasons we could discuss elsewhere.


>                    I wish to implement this system in python, making use of
> the open source CMU Sphinx system. I would love to do this under a mentor
> for gsoc 2015.
>
> Regards!
>
> Abhilash Mhaisne


See above. "voice recognition" and accent immunity is not a "python"
sort of problem, it's a massive ongoing computer and speech research
issue. Pick something *smaller*.