AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Neural Voice Cloning with a Few Samples

Sercan Arik, Jitong Chen, Kainan Peng, Wei Ping, Yanqi Zhou

NeurIPS 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Multilingual language representation pretraining · 178 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use whole-model adaptation instead of embedding-only adaptation (ablation)Speaker adaptation · VCTK (108 speakers; 10 cloning samples)MOS: 2.67$\pm$0.10 -> 3.16$\pm$0.09Improved
Use whole-model adaptation instead of embedding-only adaptation (ablation)Speaker adaptation · VCTK (108 speakers; 10 cloning samples)Similarity score: 2.95$\pm$0.09 -> 3.16$\pm$0.08Improved
Add joint fine-tuning of the encoder and generative model (ablation)Speaker encoding · VCTK (108 speakers; 10 cloning samples)MOS: 2.79$\pm$0.10 -> 2.99$\pm$0.12Improved
Add joint fine-tuning of the encoder and generative model (ablation)Speaker encoding · VCTK (108 speakers; 10 cloning samples)Similarity score: 2.85$\pm$0.10 -> 2.77$\pm$0.11Hurt
Increase the number of cloning audio samplesSpeaker adaptation and speaker encoding · VCTKNaturalness and speaker similarityImproved
Use speaker encoding rather than embedding adaptationVoice cloning · VCTKCloning time: $\sim 8$ hours -> $\sim 1.5 - 3.5$ secsImproved
Use speaker encoding instead of embedding adaptationVoice cloning · VCTKSpeaker verification performanceNo clear effect
Train speaker encoding jointly from scratch rather than separately predicting trained-model embeddings (ablation)Speaker encoding · Multi-speaker training speakersSpeaker differences in generated audioHurt
Add discriminative losses for intermediate embeddings or generated audio (ablation)Jointly trained speaker encoder · Multi-speaker training speakersSpeaker differencesNo clear effect
Add attention over cloning audio samples instead of simply averaging them (ablation)Speaker encoder · Validation set of 25 speakersEmbedding estimation errorImproved
Train the multi-speaker model on cleaner VCTK speech rather than LibriSpeechSpeaker adaptation · VCTKNaturalnessImproved
Use embedding-only rather than whole-model adaptation with limited speaker diversity (ablation)Speaker adaptation · VCTK (84 training speakers)NaturalnessHurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy