An End-to-end Chinese Text Normalization Model based on Rule-guided Flat-Lattice Transformer

TL;DR

This paper introduces an end-to-end Chinese text normalization model based on a rule-guided Flat-Lattice Transformer, effectively integrating expert rules into neural networks to improve speech synthesis clarity.

Contribution

The paper presents a novel end-to-end model that incorporates expert rules into a Flat-Lattice Transformer for Chinese text normalization, along with a large-scale dataset.

Findings

Achieved superior performance on the proposed dataset.

Effectively integrates rule-based knowledge into neural models.

Outperforms existing methods in Chinese text normalization.

Abstract

Text normalization, defined as a procedure transforming non standard words to spoken-form words, is crucial to the intelligibility of synthesized speech in text-to-speech system. Rule-based methods without considering context can not eliminate ambiguation, whereas sequence-to-sequence neural network based methods suffer from the unexpected and uninterpretable errors problem. Recently proposed hybrid system treats rule-based model and neural model as two cascaded sub-modules, where limited interaction capability makes neural network model cannot fully utilize expert knowledge contained in the rules. Inspired by Flat-LAttice Transformer (FLAT), we propose an end-to-end Chinese text normalization model, which accepts Chinese characters as direct input and integrates expert knowledge contained in rules into the neural network, both contribute to the superior performance of proposed model for the text normalization task. We also release a first publicly accessible largescale dataset for Chinese text normalization. Our proposed model has achieved excellent results on this dataset.