Accent Estimation of Japanese Words from Their Surfaces and Romanizations for Building Large Vocabulary Accent Dictionaries

TL;DR

This paper presents a method to estimate Japanese word accents from surface forms and romanizations, enabling the creation of large-scale accent dictionaries to improve TTS systems.

Contribution

The authors developed a novel accent estimation technique that predicts accents from limited information, expanding available Japanese accent dictionaries with high accuracy.

Findings

High accuracy in accent prediction for certain word categories

Successful application to NEologd dictionary

Generated large vocabulary accent dictionary with improved phonetic accuracy

Abstract

In Japanese text-to-speech (TTS), it is necessary to add accent information to the input sentence. However, there are a limited number of publicly available accent dictionaries, and those dictionaries e.g. UniDic, do not contain many compound words, proper nouns, etc., which are required in a practical TTS system. In order to build a large scale accent dictionary that contains those words, the authors developed an accent estimation technique that predicts the accent of a word from its limited information, namely the surface (e.g. kanji) and the yomi (simplified phonetic information). It is experimentally shown that the technique can estimate accents with high accuracies, especially for some categories of words. The authors applied this technique to an existing large vocabulary Japanese dictionary NEologd, and obtained a large vocabulary Japanese accent dictionary. Many cases have been observed in which the use of this dictionary yields more appropriate phonetic information than UniDic.