Strategies in Transfer Learning for Low-Resource Speech Synthesis: Phone Mapping, Features Input, and Source Language Selection
This work addresses improving speech synthesis for low-resource languages, but it is incremental as it builds on existing transfer learning methods.
The study compared phone mapping and phonological features input for transfer learning in low-resource speech synthesis, finding that both improve output quality with features performing better, but results vary by language combination. It also evaluated Angular Similarity of Phone Frequencies (ASPF) versus language distance for source language selection, showing ASPF is effective with label-based input.
We compare using a PHOIBLE-based phone mapping method and using phonological features input in transfer learning for TTS in low-resource languages. We use diverse source languages (English, Finnish, Hindi, Japanese, and Russian) and target languages (Bulgarian, Georgian, Kazakh, Swahili, Urdu, and Uzbek) to test the language-independence of the methods and enhance the findings' applicability. We use Character Error Rates from automatic speech recognition and predicted Mean Opinion Scores for evaluation. Results show that both phone mapping and features input improve the output quality and the latter performs better, but these effects also depend on the specific language combination. We also compare the recently-proposed Angular Similarity of Phone Frequencies (ASPF) with a family tree-based distance measure as a criterion to select source languages in transfer learning. ASPF proves effective if label-based phone input is used, while the language distance does not have expected effects.