Voice-preserving Zero-shot Multiple Accent Conversion
This addresses the difficulty in communication and language learning by enabling zero-shot multiple accent conversion without changing voice identity, though it is incremental as it builds on existing accent conversion models.
The paper tackled the problem of accent conversion while preserving speaker identity by using adversarial learning to disentangle accent-dependent features, achieving audio that sounds closer to the target accent and like the original speaker in subjective evaluations.
Most people who have tried to learn a foreign language would have experienced difficulties understanding or speaking with a native speaker's accent. For native speakers, understanding or speaking a new accent is likewise a difficult task. An accent conversion system that changes a speaker's accent but preserves that speaker's voice identity, such as timbre and pitch, has the potential for a range of applications, such as communication, language learning, and entertainment. Existing accent conversion models tend to change the speaker identity and accent at the same time. Here, we use adversarial learning to disentangle accent dependent features while retaining other acoustic characteristics. What sets our work apart from existing accent conversion models is the capability to convert an unseen speaker's utterance to multiple accents while preserving its original voice identity. Subjective evaluations show that our model generates audio that sound closer to the target accent and like the original speaker.