Published Version
Open DOI LinkAbstract
Recent works on voice conversion (VC) focus on preserving the rhythm and the intonation as well as the linguistic content. To preserve these features from the source, we decompose current non-parallel VC systems into two encoders and one decoder. We analyze each module with several experiments and reassemble the best components to propose Assem-VC, a new state-of-the-art any-to-many non-parallel VC system. We also examine that PPG and Cotatron features are speaker-dependent, and attempt to remove speaker identity with adversarial training. Code and audio samples are available at https://github.com/mindslab-ai/assem-vc.
Full Text
Open DOI Link
Sign-in/Register to access full text options Topics from this Paper
Voice Conversion- Linguistic Content
- Adversarial Training
- Audio Samples
- PPG Features + Show 5 more
Create a personalized feed of these topics
Get StartedSimilar Papers
VQVC+: One-Shot Voice Conversion by Vector Quantization and U-Net Architecture 
Relevant
Not Relevant
Oct 25, 2020
RelevantNot Relevant A Survey on Generative Adversarial Networks based Models for Many-to-many Non-parallel Voice Conversion RelevantNot Relevant
- Mar 9, 2022
RelevantNot Relevant Many-To-Many Voice Conversion Using Conditional Cycle-Consistent Adversarial Networks RelevantNot Relevant
- May 1, 2020
RelevantNot Relevant Many-to-Many Voice Conversion Based Feature Disentanglement Using Variational Autoencoder RelevantNot Relevant
- Aug 30, 2021
RelevantNot Relevant Crank: An Open-Source Software for Nonparallel Voice Conversion Based on Vector-Quantized Variational Autoencoder RelevantNot Relevant
- Jun 6, 2021
RelevantNot Relevant Voice Conversion of Tagalog Synthesized Speech Using Cycle-Generative Adversarial Networks (Cycle-GAN) RelevantNot Relevant
- Oct 21, 2022
RelevantNot Relevant CycleGAN-VC3: Examining and Improving CycleGAN-VCs for Mel-Spectrogram Conversion RelevantNot Relevant
- Oct 25, 2020
RelevantNot Relevant Transfer Learning From Speech Synthesis to Voice Conversion With Non-Parallel Training Data RelevantNot Relevant
IEEE Transactions on Audio, Speech, and Language Processing- Jan 1, 2021
RelevantNot Relevant Transfer Learning From Speech Synthesis to Voice Conversion With Non-Parallel Training Data RelevantNot Relevant
- IEEE/ACM Transactions on Audio, Speech, and Language Processing
- Jan 1, 2021
RelevantNot Relevant Emotional Voice Conversion Using Multitask Learning with Text-To-Speech RelevantNot Relevant
- May 1, 2020
RelevantNot Relevant Unsupervised End-to-End Learning of Discrete Linguistic Units for Voice Conversion RelevantNot Relevant
- Sep 15, 2019
RelevantNot Relevant Non-parallel Voice Conversion with Controllable Speaker Individuality using Variational Autoencoder RelevantNot Relevant
- Nov 1, 2019
RelevantNot Relevant On the Study of Generative Adversarial Networks for Cross-Lingual Voice Conversion RelevantNot Relevant
- Dec 1, 2019
RelevantNot Relevant