Back-Translation-Style Data Augmentation for Mandarin Chinese Polyphone Disambiguation
This addresses a key bottleneck in Mandarin TTS systems by augmenting scarce annotated datasets, though it is an incremental application of existing techniques to a specific domain.
The paper tackles the problem of Mandarin Chinese polyphone disambiguation for Text-To-Speech systems by proposing a back-translation-style data augmentation method using unlabeled text data, which improves accuracy by leveraging pseudo-labels and data balancing strategies.
Conversion of Chinese Grapheme-to-Phoneme (G2P) plays an important role in Mandarin Chinese Text-To-Speech (TTS) systems, where one of the biggest challenges is the task of polyphone disambiguation. Most of the previous polyphone disambiguation models are trained on manually annotated datasets, and publicly available datasets for polyphone disambiguation are scarce. In this paper we propose a simple back-translation-style data augmentation method for mandarin Chinese polyphone disambiguation, utilizing a large amount of unlabeled text data. Inspired by the back-translation technique proposed in the field of machine translation, we build a Grapheme-to-Phoneme (G2P) model to predict the pronunciation of polyphonic character, and a Phoneme-to-Grapheme (P2G) model to predict pronunciation into text. Meanwhile, a window-based matching strategy and a multi-model scoring strategy are proposed to judge the correctness of the pseudo-label. We design a data balance strategy to improve the accuracy of some typical polyphonic characters in the training set with imbalanced distribution or data scarcity. The experimental result shows the effectiveness of the proposed back-translation-style data augmentation method.