Our objective is to demonstrate the applicability of adaptive wavelets
for speech applications. In particular, we discuss two applications,
namely, classification of unvoiced sounds and speaker identification.
First, a method to classify unvoiced sounds using adaptive wavelets, w
hich would help in developing a unified algorithm to classify phonemes
(speech sounds), is described. Next, the applicability of adaptive wa
velets to identify speakers using very short speech data (one pitch pe
riod) is exhibited. The described text-independent phoneme based speak
er identification algorithm identifies a speaker by first modeling pho
nemes and then by clustering all the phonemes belonging to the same sp
eaker into one class. For both applications, we use feed-forward neura
l network architecture. We demonstrate the performance of both unvoice
d sounds classifier and speaker identification algorithms by using rep
resentative real speech examples.