metadata
license: cc-by-nc-4.0
pipeline_tag: audio-to-audio
library_name: f5-tts
extra_gated_prompt: >-
You agree to not use the model to generate, share, or promote content that is
illegal, harmful, deceptive, or intended to impersonate real individuals
without their informed consent.
extra_gated_fields:
Affiliation: text
Country: country
I agree to use this model for non-commercial use ONLY: checkbox
EZ-VC: Easy Zero-shot Any-to-Any Voice Conversion
Our paper has been published in the Findings of EMNLP 2025!
Installation
Create a separate environment if needed
# Create a python 3.10 conda env (you could also use virtualenv)
conda create -n ez-vc python=3.10
conda activate ez-vc
Local installation
git clone https://github.com/EZ-VC/EZ-VC
cd EZ-VC
git submodule update --init --recursive
pip install -e .
# Install espnet for xeus (Exactly this version)
pip install 'espnet @ git+https://github.com/wanchichen/espnet.git@ssl'
Inference
We have provided a Jupyter notebook for inference in "src/f5_tts/infer/infer.ipynb".
Open Inference notebook.
Run all.
The converted audio will be available at the last cell.
Acknowledgements
- F5-TTS for opensourcing their code which has made EZ-VC possible.
Citation
If our work and codebase is useful for you, please cite as:
@inproceedings{joglekar-etal-2025-ez,
title = "{EZ}-{VC}: Easy Zero-shot Any-to-Any Voice Conversion",
author = "Joglekar, Advait and
Singh, Divyanshu and
Bhatia, Rooshil Rohit and
Umesh, Srinivasan",
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-emnlp.1077/",
doi = "10.18653/v1/2025.findings-emnlp.1077",
pages = "19768--19774",
ISBN = "979-8-89176-335-7",
abstract = "Voice Conversion research in recent times has increasingly focused on improving the zero-shot capabilities of existing methods. Despite remarkable advancements, current architectures still tend to struggle in zero-shot cross-lingual settings. They are also often unable to generalize for speakers of unseen languages and accents. In this paper, we adopt a simple yet effective approach that combines discrete speech representations from self-supervised models with a non-autoregressive Diffusion-Transformer based conditional flow matching speech decoder. We show that this architecture allows us to train a voice-conversion model in a purely textless, self-supervised fashion. Our technique works without requiring multiple encoders to disentangle speech features. Our model also manages to excel in zero-shot cross-lingual settings even for unseen languages. We provide our code, model checkpoint and demo samples here: https://github.com/ez-vc/ez-vc"
}
License
Our code is released under MIT License. The pre-trained models are licensed under the CC-BY-NC license. Sorry for any inconvenience this may cause.