This research presents a novel method for converting the sound of a speaker using a two-stage process: a training stage and a conversion stage. In the training stage, fundamental frequency, speaker, and content characteristics are extracted from training voice signals of both a source speaker and a target speaker. A fundamental frequency conversion function and a speaker conversion function are constructed based on these characteristics. In the conversion stage, the fundamental frequency and spectrum characteristics are extracted from the voice signal to be converted from the source speaker. The fundamental frequency and speaker conversion functions obtained from the training stage are then applied to convert the extracted characteristics. The converted fundamental frequency and speaker characteristics are used to synthesize voices of the target speaker in the converted voice signal. The method offers ease of implementation and achieves superior sound quality and similarity in the converted output.

Speaker conversion, sound conversion, fundamental frequency, speaker characteristic, content characteristic, voice synthesis.