Towards Building Text-to-Speech Systems for the Next Billion Users
Document Type
Conference Proceeding
Publication Title
ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Abstract
Deep learning based text-to-speech (TTS) systems have been evolving rapidly with advances in model architectures, training methodologies, and generalization across speakers and languages. However, these advances have not been thoroughly investigated for Indian language speech synthesis. Such investigation is computationally expensive given the number and diversity of Indian languages, relatively lower resource availability, and the diverse set of advances in neural TTS that remain untested. In this paper, we evaluate the choice of acoustic models, vocoders, supplementary loss functions, training schedules, and speaker and language diversity for Dravidian and Indo-Aryan languages. Based on this, we identify monolingual models with FastPitch and HiFi-GAN V1, trained jointly on male and female speakers to perform the best. With this setup, we train and evaluate TTS models for 13 languages and find our models to significantly improve upon existing models in all languages as measured by mean opinion scores. We open-source all models on the Bhashini platform.
DOI
10.1109/ICASSP49357.2023.10096069
Publication Date
1-1-2023
Keywords
indian languages, text-to-speech
Recommended Citation
G. Kumar et al., "Towards Building Text-to-Speech Systems for the Next Billion Users," ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, vol. 2023-June, Jan 2023.
The definitive version is available at https://doi.org/10.1109/ICASSP49357.2023.10096069