Implementation of an Voice Gender Converter
Tony’s research is focused on implementation of DSP
algorithms into hardware using high level VLSI design. At present Tony is
working on a Voice Gender Normalisation algorithm to be implemented into CMOS
technology. This will result in changing the gender speech characteristics of a
speaker so their speech sounds like a member of the opposite sex, i.e., male to
female or female to male conversion. The technology will be used in the area of
speech compression for transmission purposes, speech recognition and voice
disguise. The algorithm will be implemented in real time to enable use in mobile
phones and other communication devices.
Tony’s most recent project was implementing a real
time pitch detector into a CMOS technology. This was used in conjunction with an
adaptive filter to achieve real time noise cancellation in a speech signal.