2013Journal of Tsinghua University(Science and Technology)Requires access

Speech synthesis for Chinese dialects based on the phonological system of ancient Chinese

Ziyu Xiong

Open publisher page 0 citations

Abstract

The HTS speech training synthesis tools and STRAIGHT speech synthesizer are used for speech training synthesis of an unknown dialect phonetic system.This study applies these methods to various Chinese dialects by designing a set of universal pronunciation texts for Chinese dialect speech synthesis with recording of a synthesized speech corpus for some Chinese dialects.A Chinese dialect speech synthesis platform is then built based on the phonological system of ancient Chinese.Results for Mandarin speech synthesis show that the synthesized sound based on the phonological system of ancient Chinese is quite close in terms of intelligibility and quality to the synthesized sound based on the pinyin system.Thus,the speech synthesis based on the phonological system of ancient Chinese for Chinese dialects is effective and feasible.

About this research paper

What this paper is about

The HTS speech training synthesis tools and STRAIGHT speech synthesizer are used for speech training synthesis of an unknown dialect phonetic system.This study applies these methods to various Chinese dialects by designing a set of universal pronunciation texts for Chinese dialect speech synthesis with recording of a synthesized speech corpus for some Chinese dialects.A Chinese dialect speech synthesis platform is then built based on the phonological system of ancient Chinese.Results for Mandarin speech synthesis show that the synthesized sound based on the phonological system of ancient Chinese is quite close in terms of intelligibility and quality to the synthesized sound based on the pinyin system.Thus,the speech synthesis based on the phonological system of ancient Chinese for Chinese dialects is effective and feasible.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

The HTS speech training synthesis tools and STRAIGHT speech synthesizer are used for speech training synthesis of an unknown dialect phonetic system.This study applies these methods to various Chinese dialects by designing a set of universal pronunciation texts for Chinese dialect speech synthesis with recording of a synthesized speech corpus for some Chinese dialects.A Chinese dialect speech synthesis platform is then built based on the phonological system of ancient Chinese.Results for Mandarin speech synthesis show that the synthesized sound based on the phonological system of ancient Chinese is quite close in terms of intelligibility and quality to the synthesized sound based on the pinyin system.Thus,the speech synthesis based on the phonological system of ancient Chinese for Chinese dialects is effective and feasible.

Key concepts: Chinese speech synthesis, Speech synthesis, Mandarin Chinese, Pinyin, Pronunciation, Computer science, Speech corpus, Speech recognition

Related papers

Back to paper searchBrowse research topicsOriginal source
Speech synthesis for Chinese dialects based on the phonological system of ancient Chinese — Research Paper | ScholarLens