De l'utilisation de descripteurs issus de la linguistique computationnelle dans le cadre de la synthèse par HMM

Abstract : Over the last decades, acoustic modeling for speech synthesis has been improved significantly. However, in most systems, the descriptive feature set used to represent annotated text has been the same for many years. Specifically, the prosody models in most systems are based on low level information such as syllable stress or word part-of-speech tags. In this paper, we propose to enrich the descriptive feature set by adding a linguistic measure computed from the predictability of an event, such as the occurrence of a syllable or word. By adding such descriptive features, we assume that we will improve prosody modeling. This new feature set is then used to train prosody models for speech synthesis. This paper focuses on an objective analysis of the influence of these descriptive features on the synthesis achieved in English and French.
Complete list of metadatas

https://hal.inria.fr/hal-01338953
Contributor : Damien Lolive <>
Submitted on : Wednesday, June 29, 2016 - 1:41:35 PM
Last modification on : Wednesday, January 23, 2019 - 11:26:02 AM

Identifiers

  • HAL Id : hal-01338953, version 1

Citation

Bernd Möbius, Sébastien Le Maguer, Ingmar Steiner, Damien Lolive. De l'utilisation de descripteurs issus de la linguistique computationnelle dans le cadre de la synthèse par HMM. Journées d'Études sur la Parole, Jul 2016, Paris, France. ⟨hal-01338953⟩

Share

Metrics

Record views

259