[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Neural Sound Processing Schemes

I imagine that the most important signal processing you are after isn't
signal processing at all in the classical sense, in that the 'top-down'
components are so ubiquitous and unquantifiable. That is, recognition of
different pronouncements / acoustic conditions relies on 'context
processing' and hypothesis testing far more than a strict analysis of audio
signal data.
----- Original Message -----
From: "M & E" <mariusm@PARYS.CO.ZA>
Sent: 26 March 2002 18:52
Subject: Neural Sound Processing Schemes

> Hi,
> I am developing a very advanced artificial intelligence system.  I am
> currently thinking about implementing sound recognition in the system.  I
> have already coded a .wav reading function, since sounds will be presented
> in this format.  Now I need some suggestions with the actual processing of
> the data.  Any suggestions will be much appreciated.
> The system is very much based on neurobiology, and almost any neurally
> plausible organization can be implemented in my AI system.  I am now
> investigating some of the important mechanisms employed by the ear and
> to process auditory data.  What kinds of signal processing is done to
> uniquely identify different sounds?  And specifically, how are different
> pronouncements of the same word, recognized as the same word by the brain,
> even if the actual audio data is very different?
> Thank you,
> Marius Myburg.
> http://farve.united.net.kg