Sorry, you need to enable JavaScript to visit this website.

DISCOURSE-LEVEL PROSODY MODELING WITH A VARIATIONAL AUTOENCODER FOR NON-AUTOREGRESSIVE EXPRESSIVE SPEECH SYNTHESIS

Citation Author(s):
Ning-Qian wu, Zhao-Ci Liu, Zhen-Hua Ling
Submitted by:
Ningqian Wu
Last updated:
10 May 2022 - 11:18am
Document Type:
Presentation Slides
Document Year:
2022
Event:
Presenters:
Ning-Qian Wu
Paper Code:
SPE-55.4

Abstract

To address the issue of one-to-many mapping from phoneme sequences to acoustic features in expressive speech synthesis, this paper proposes a method of discourse-level prosody modeling with a variational autoencoder (VAE) based on the non-autoregressive architecture of FastSpeech. In this method, phone-level prosody codes are extracted from prosody features by combining VAE with FastSpeech, and are predicted using discourse-level text features together with BERT embeddings. The continuous wavelet transform (CWT) in FastSpeech2 for F0 representation is not necessary anymore. Experimental results on a Chinese audiobook dataset show that our proposed method can effectively take advantage of discourse-level linguistic information and has outperformed FastSpeech2 on the naturalness and expressiveness of synthetic speech.

up
0 users have voted:

Files

presentation slides

(41)