THE IDENTIFICATION OF GENDER FROM SYNTHETIC SPEECH
SP WHITESIDE, J GOODALL · 2024
this paper presents ndings on some of the acoustic cues that listeners use to make decisions about the gender of an adult W from synthetic speech.Four young adult male and adult female speakers from South orlrshire.Fmgland were chosen for the irutial part of this study.They were selected on the basis of criteria which Included age.accent and normal 3 hearing Subsequently recordings of the eight speakers were pla listeners.The were asked to rate the gender of the n the basis of this one male and one fem e were selected to provide data for analysis.synthesis and perce tion test procedures.ustic Eammeters of the male speaker's data were obtained using a Y Computerised Speech La (CSL) model 4300.These parameters were used as reference values to manipulate the female s r's acoustic patterns using an LPC parameter mani ulationlsymhesis program (model 4304, version 2.01) with an editing facility.The fem e s 's v0wels were edited with either a lowered fundamental frequenc and/or