Hi,
..
I have a pretty unreadable printed copy of above paper from M.K.Alisd...,
M.S. Candidate at Columbia University from Spring 2002. It has a pretty
simple description of use of TDOA approach in separating two speakers from
ShATR database and coresponding Matlab scripts. As said, it's hardly
readable, so I'm searching fro electronic version (it points to
http://www.columbia.edu/~mka2003/speech.html
, but this is nonexistant ).
I'd kindly ask if anyone has that electronic copy and is willing to share.
I'd like to use it to create lab exercise for students...
Also any other pointers to similar info will be much appreciated... Are
some
samples from ShATR corpus available somewhere ?
Thanks in advace,
regards,
Bulek.