Continuous probabilistic transform for voice conversion

TLDR

Voice conversion modifies a source speaker’s speech to sound as if spoken by a target speaker. The paper proposes a new method to model the relationship between source and target spectral envelopes. The method employs a Gaussian mixture model of source spectral envelopes and represents conversion as a continuous parametric function whose parameters are estimated by least‑squares optimization and implemented within an HNM system. Compared with vector‑quantization approaches, the transform achieves a closer match to target envelopes and, according to objective and listening tests, markedly improves the quality and naturalness of the converted speech.

Abstract

Voice conversion, as considered in this paper, is defined as modifying the speech signal of one speaker (source speaker) so that it sounds as if it had been pronounced by a different speaker (target speaker). Our contribution includes the design of a new methodology for representing the relationship between two sets of spectral envelopes. The proposed method is based on the use of a Gaussian mixture model of the source speaker spectral envelopes. The conversion itself is represented by a continuous parametric function which takes into account the probabilistic classification provided by the mixture model. The parameters of the conversion function are estimated by least squares optimization on the training data. This conversion method is implemented in the context of the HNM (harmonic+noise model) system, which allows high-quality modifications of speech signals. Compared to earlier methods based on vector quantization, the proposed conversion scheme results in a much better match between the converted envelopes and the target envelopes. Evaluation by objective tests and formal listening tests shows that the proposed transform greatly improves the quality and naturalness of the converted speech signals compared with previous proposed conversion methods.