Third International Conference on Spoken Language Processing (ICSLP 94)

Yokohama, Japan
September 18-22, 1994

Speaking Style Conversion by Changing Prosodic Parameters and Formant Frequencies

Masanobu Abe, Hideyuki Mizuno

NTT Human Interface Labs., Kanagawa, Japan

The aim of this paper is to generate various speaking styles with the goal of enhancing the speech uttered by a person. Speaking style conversion is achieved by applying methods that we have proposed for text-to-speech systems and voice conversion. As the first trial, we selected three speaking styles appropriate for task specific texts, and analyzed their characteristics. Large differences were found in 1st and 3rd formant frequency, F0range and phrase height assignments, speaking rate, and segmental duration of phrases followed by pauses. Preliminary experiments of speaking style conversion are performed. Listening tests show that the proposed approach can freely change the speaking style.

Full Paper

Bibliographic reference.  Abe, Masanobu / Mizuno, Hideyuki (1994): "Speaking style conversion by changing prosodic parameters and formant frequencies", In ICSLP-1994, 1455-1458.