Hi All
I'm doing on Robust Speaker identification/verification for telephony
applications for my final year project.I pass through my clean speech signal to
low-pass filter with cut-off frequency 3.4kHz. I want to use inverse filtering
about 4~5kHz.
I think the step will be
1. transform the telephony speech into frequency domain
2. i need to get back the original S(z)=c(z)/H(z)
where S(z) is original speech ,c(z)=telephony speech ,H(z)=inverse filter
how to find inverse filter transfer function and is the step correct?
Pls guide to me
with regards
