.png){kind=icon}
.png){kind=icon}
Return
Pretraining Techniques for Sequence-to-Sequence Voice Conversion
.png){kind=icon}
OA
AI
DOI：10.1109/TASLP.2021.3049336.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Sequence-to-sequence (seq2seq) voice conversion (VC) models are attractive owing to their ability to convert prosody. Nonetheless, without sufficient data, seq2seq VC models can suffer from unstable training and mispronunciation problems in the converted speech, thus far from practical. To tackle these shortcomings, we propose to transfer knowledge from other speech processing tasks where large-scale corpora are easily available, typically text-to-speech (TTS) and automatic speech recognition (ASR). We argue that VC models initialized with such pretrained ASR or TTS model parameters can generate effective hidden representations for high-fidelity, highly intelligible converted speech. In this work, we examine our proposed method in a parallel, one-to-one setting. We employed recurrent neural network (RNN)-based and Transformer based models, and through systematical experiments, we demonstrate the effectiveness of the pretraining scheme and the superiority of Transformer based models over RNN-based models in terms of intelligibility, naturalness, and similarity.
Keywords:
Task analysis
Speech processing
Decoding
Training
Data models
Training data
Spectrogram
Voice conversion
sequence-to-sequence
pretraining
transformer
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
I
IF:
5.1
Papers:
2.6K
Citations:
1.1W