Cross-speaker Emotion Transfer by Manipulating Speech Style Latents
This addresses the challenge of limited labeled data and poor controllability in emotional text-to-speech for speech synthesis applications, though it is incremental.
The paper tackles the problem of generating emotional speech without requiring large labeled datasets and enables control over emotion intensity, achieving superior performance in expressiveness, naturalness, and controllability while preserving speaker identity.
In recent years, emotional text-to-speech has shown considerable progress. However, it requires a large amount of labeled data, which is not easily accessible. Even if it is possible to acquire an emotional speech dataset, there is still a limitation in controlling emotion intensity. In this work, we propose a novel method for cross-speaker emotion transfer and manipulation using vector arithmetic in latent style space. By leveraging only a few labeled samples, we generate emotional speech from reading-style speech without losing the speaker identity. Furthermore, emotion strength is readily controllable using a scalar value, providing an intuitive way for users to manipulate speech. Experimental results show the proposed method affords superior performance in terms of expressiveness, naturalness, and controllability, preserving speaker identity.