Hiding speaker's sex in speech using zero-evidence speaker representation in an analysis/synthesis pipeline

The use of modern vocoders in an analysis/synthesis pipeline allows to investigate high-quality voice conversion which can be used for privacy purposes. Here, we proposed to transform the speaker embedding and the pitch in order to hide the sex of the speaker. The ECAPA-TDNN based speaker representation fed into a HifiGAN vocoder is protected using a neural-discriminant analysis approach which is consistent with the zero-evidence concept of privacy. This approach significantly reduces in the speech the information related to the speaker’s sex while preserving speech content and some consistency in the resulting protected voices.