Fast-VGAN: Lightweight Voice Conversion with Explicit Control of F0 and Duration Parameters
This work addresses the problem of flexible voice transformation for applications in speech synthesis and conversion, though it appears incremental as it builds on existing neural vocoder methods.
The paper tackled the challenge of precise control over speech characteristics like pitch and duration in voice conversion by proposing a convolutional neural network-based model explicitly conditioned on factors such as fundamental frequency and phoneme sequences to generate mel spectrograms, achieving high intelligibility and speaker similarity in evaluations.
Precise control over speech characteristics, such as pitch, duration, and speech rate, remains a significant challenge in the field of voice conversion. The ability to manipulate parameters like pitch and syllable rate is an important element for effective identity conversion, but can also be used independently for voice transformation, achieving goals that were historically addressed by vocoder-based methods. In this work, we explore a convolutional neural network-based approach that aims to provide means for modifying fundamental frequency (F0), phoneme sequences, intensity, and speaker identity. Rather than relying on disentanglement techniques, our model is explicitly conditioned on these factors to generate mel spectrograms, which are then converted into waveforms using a universal neural vocoder. Accordingly, during inference, F0 contours, phoneme sequences, and speaker embeddings can be freely adjusted, allowing for intuitively controlled voice transformations. We evaluate our approach on speaker conversion and expressive speech tasks using both perceptual and objective metrics. The results suggest that the proposed method offers substantial flexibility, while maintaining high intelligibility and speaker similarity.