mlsanityData to 14 Oct 2026

Strategies in Transfer Learning for Low-Resource Speech Synthesis: Phone Mapping, Features Input, and Source Language Selection

arXiv 2306.12040 ↗First posted 21 Jun 2023cs.CL eess.AScited by 2

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 19 of its 35 references that are in the map. First posted on arXiv 21 Jun 2023.

Builds on

19 of 35 references.

References · 19

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,905
  3. 2019Common Voice: A Massively-Multilingual Speech Corpuscited by 1,384
  4. 2020HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesiscited by 1,154
  5. 2017Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictionscited by 804
  6. 2020FastSpeech 2: Fast and High-Quality End-to-End Text to Speechcited by 717
  7. 2021A Survey on Neural Speech Synthesiscited by 184
  8. 2019MOSNet: Deep Learning based Objective Assessment for Voice Conversioncited by 127
  9. 2021Generalization Ability of MOS Prediction Networkscited by 69
  10. 2022The VoiceMOS Challenge 2022cited by 57
  11. 2019CSS10: A Collection of Single Speaker Speech Datasets for 10 Languagescited by 54
  12. 2019Multilingual Neural Machine Translation with Language Clusteringcited by 47
7 more
  1. 2021Investigating on Incorporating Pretrained and Learnable Speaker Representations for Multi-Speaker Multi-Style Text-to-Speechcited by 37
  2. 2019End-to-end Text-to-speech for Low-resource Languages by Cross-Lingual Transfer Learningcited by 30
  3. 2021How do Voices from Past Speech Synthesis Challenges Compare Today?cited by 21
  4. 2020Phonological Features for 0-shot Multilingual Speech Synthesiscited by 20
  5. 2022Language-Agnostic Meta-Learning for Low-Resource Text-to-Speech with Articulatory Featurescited by 20
  6. 2022SOMOS: The Samsung Open MOS Dataset for the Evaluation of Neural Text-to-Speech Synthesiscited by 18
  7. 2023Resource-Efficient Fine-Tuning Strategies for Automatic MOS Prediction in Text-to-Speech for Low-Resource Languagescited by 3

Built on by

Cited by year

Citing papers per year: 2024: 2.

2 papers.

  1. 2024Meta Learning Text-to-Speech Synthesis in over 7000 Languagescited by 9
  2. 2024An Initial Investigation of Language Adaptation for TTS Systems under Low-resource Scenarioscited by 4