Chatr : a multi-lingual speech re-sequencing synthesis system

CAMPBELL Nick, BLACK Alan W.

Bibliographic Information

Other Title

CHATR : 自然音声波形接続型任意音声合成システム

Description

This paper describes a method for producing speech synthesis without signal processing, using re-sequencing of phone-sized segmets from a pre-recorded speech corpus for the purpose of reproducing the voice characteristics and speaking style of the original speaker to create novel utterances. We describe procedures for indexing and retrieval that make the synthesiser independent of language or speaker. A re-sequencing speech synthesiser doesn't produce speech sounds; it produces an index for a random-access retrieval sequence from the original speech to give the closest approximation to a desired specification from the segments available in a given speech corpus. To find the optimal sequence of segments for concatenation, the synthesiser first creates an inventory of phones and their acoustical characteristics, and then selects from amongst these by a weighted combination of the features to give an index of the segment sequence that best matches the target specification.

Journal

IEICE technical report. Speech

IEICE technical report. Speech 96 (39), 45-52, 1996-05-16

The Institute of Electronics, Information and Communication Engineers

Keywords

Details 詳細情報について

CRID: 1570291227409416832

NII Article ID: 110003296229

NII Book ID: AN10013221

Text Lang: ja

Data Source

CiNii Articles

Export

Report a problem