Investigating On Incorporating Pretrained And Learnable Speaker Representations For Multi-speaker Multi-style Text-to-speech

Abstract

The few-shot multi-speaker multi-style voice cloning task is to synthesize utterances with voice and speaking style similar to a reference speaker given only a few reference samples. In this work, we investigate different speaker representations and proposed to integrate pretrained and learnable speaker representations. Among different types of embeddings, the embedding pretrained by voice conversion achieves the best performance. The FastSpeech 2 model combined with both pretrained and learnable speaker representations shows great generalization ability on few-shot speakers and achieved 2nd place in the one-shot track of the ICASSP 2021 M2VoC challenge.

Stats

• citations35
• S2 citations—
• github stars0
• HF likes0
• heat score11.67
• arxiv keychien2021investigating

Related papers

• Meta-voice: Fast Few-shot Style Transfer For Expressive Voice Cloning Using Meta Learning (2021)0.00
• The THU-HCSI Multi-speaker Multi-lingual Few-shot Voice Cloning System For LIMMITS'24 Challenge (2024)0.00
• Unet-tts: Improving Unseen Speaker And Style Transfer In One-shot Voice Cloning (2021)0.00
• Voice Cloning: A Multi-speaker Text-to-speech Synthesis Approach Based On Transfer Learning (2021)0.00
• U-style: Cascading U-nets With Multi-level Speaker And Style Modeling For Zero-shot Voice Cloning (2023)6.77
• One-shot Voice Conversion For Style Transfer Based On Speaker Adaptation (2021)8.09
• Single And Multi-speaker Cloned Voice Detection: From Perceptual To Learned Features (2023)9.23
• Controlspeech: Towards Simultaneous And Independent Zero-shot Speaker Cloning And Zero-shot Language Style Control (2024)9.40