Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
HQ-SVC: Towards High-Quality Zero-Shot Singing Voice Conversion in Low-Resource Scenarios
Bingsong Bai, Yizhong Geng, Fengping Wang, Cong Wang, Puyuan Guo, Yingming Gao, Ya Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add diffusion model after DDSP (ablation) | HQ-SVC · Zero-shot SVC | SECS: 0.420 -> 0.627 | Improved |
| add diffusion model after DDSP (ablation) | HQ-SVC · Zero-shot SVC | NISQA: 3.175 -> 3.841 | Improved |
| add diffusion model after DDSP (ablation) | HQ-SVC · Zero-shot SVC | STOI: 0.820 -> 0.799 | Hurt |
| add DDSP module (ablation) | HQ-SVC · Zero-shot SVC | SECS: 0.608 -> 0.627 | Improved |
| add DDSP module (ablation) | HQ-SVC · Zero-shot SVC | STOI and NISQA: 0.847 -> 0.799 STOI; 3.960 -> 3.841 NISQA | Hurt |
| use FACodec single decoupled encoder instead of separate CAM++ and ContentVec (ablation) | HQ-SVC · Zero-shot SVC | SECS: 0.668 -> 0.627 | Hurt |
| use FACodec single decoupled encoder instead of separate CAM++ and ContentVec (ablation) | HQ-SVC · Zero-shot SVC | STOI and F0 RMSE: STOI 0.738 -> 0.799; F0 RMSE 9.803 -> 8.681 | Improved |
| add speaker loss Lspk (ablation) | HQ-SVC · Zero-shot SVC | SECS: 0.622 -> 0.627 | Improved |
| add speaker loss Lspk (ablation) | HQ-SVC · Zero-shot SVC | SMOS: 3.688 -> 3.578 | Hurt |
| add Speaker-F0 predictor loss Lf0 (ablation) | HQ-SVC · Zero-shot SVC | F0 RMSE: 8.855 -> 8.681 | Improved |
| add Lf0 (ablation) | HQ-SVC · Zero-shot SVC | SMOS: 3.729 -> 3.578 | Hurt |
| use HQ-SVC instead of SaMoye-SVC | zero-shot SVC · Zero-shot SVC | SECS: 0.647 -> 0.627 | Hurt |
Closest papers by meaning
- StableVC: Style Controllable Zero-Shot Voice Conversion with Conditional Flow Matching (AAAI 2025 · similarity 0.89)
- NaturalSpeech 2: Latent Diffusion Models are Natural and Zero-Shot Speech and Singing Synthesizers (ICLR 2024 · similarity 0.86)
- Vevo: Controllable Zero-Shot Voice Imitation with Self-Supervised Disentanglement (ICLR 2025 · similarity 0.86)
- StyleSinger: Style Transfer for Out-of-Domain Singing Voice Synthesis (AAAI 2024 · similarity 0.86)
- CSSinger: End-to-End Chunkwise Streaming Singing Voice Synthesis System Based on Conditional Variational Autoencoder (AAAI 2025 · similarity 0.85)
- Phoneme Hallucinator: One-Shot Voice Conversion via Set Expansion (AAAI 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
