Paraphrase generation to improve text-to-speech synthesis

Text-to-speech synthesizer systems are of overall good quality, especially when adapted to a specific task. Given this task and an adapted voice corpus, the message quality is mainly dependent on the wording used. This paper presents how a paraphrase generator can be used in synergy with a text-to-speech (TTS) synthesis system to improve its overall performances. Our system is composed of a paraphrase generator using a French-toFrench corpus learnt on a bilingual aligned corpus, a TTS selector based on the unit selection cost, and a TTS synthesizer. We present an evaluation of the system, which highlights the need for systematic subjective evaluation.

• Type: Journal
• Year: 2010
• Where: INTERSPEECH
• Authors: Ghislain Putois, Jonathan Chevelu, Cédric Boidin
• Keywords: Adapted Voice Corpus, INTERSPEECH 2010, Paraphrase Generator, Signal Processing, Text-to-speech Synthesizer Systems