May 21, 1981, 3:21:18 PM5/21/81
Is there anybody out there (perhaps somebody at Bell Labs?) working on
pattern recognition etc. as applied to voiceprints? No, we aren't working
for the Defence department: we have people using sound spectrograms for
analysis of things like birdsongs. But right now, all the analysis is
pretty much manual, and there is some interest in automating some of it
at some time. Anybody doing anything along these lines?