Language-Agnostic Meta-Learning for Low-Resource Text-to-Speech with Articulatory Features
This enables text-to-speech for many of the over 6,000 spoken languages with limited data, addressing a significant gap in accessibility.
The paper tackles the problem of low-resource text-to-speech for languages lacking training data by using articulatory features and meta-learning, achieving high-quality synthesis with only 30 minutes of data in an unseen language and speaker.
While neural text-to-speech systems perform remarkably well in high-resource scenarios, they cannot be applied to the majority of the over 6,000 spoken languages in the world due to a lack of appropriate training data. In this work, we use embeddings derived from articulatory vectors rather than embeddings derived from phoneme identities to learn phoneme representations that hold across languages. In conjunction with language agnostic meta learning, this enables us to fine-tune a high-quality text-to-speech model on just 30 minutes of data in a previously unseen language spoken by a previously unseen speaker.