LiteVITS Data

This repository contains the data used in the paper:

“A Simple and Affordable Adversarial Attack By Speech Generation Finetuning”

• authors : Elie Laurent Benaroya, Damien Lolive, Arnaud Delhay, Hoan My Tran
• accepted at 2026 IEEE Spoken Language Technology Workshop (SLT 2026)

The data support the experiments presented in the paper and are intended for research on speech generation and adversarial attacks against anti-spoofing countermeasure models.

The corresponding code and training framework are available at:

https://gitlab.inria.fr/expression/litevits.git

Please refer to the paper (when published) and the code repository for details about the data preparation, experimental setup, and usage.