Skip to content

Dataset · Wan-fen52

MAVOS-DD multilingual deepfake detection benchmark

Wan-fen52/MAVOS-DD

An open multilingual audio-video dataset for deepfake detection.

What's inside

Contains train, validation, and open-set splits for researching deepfake detectors. It supports multiple languages (ar, ro, en, de, hi, es) and is delivered in the Hugging Face Dataset format.

What you can do with it

  • Train and evaluate audio-video deepfake detection models
  • Test the robustness of open-set systems
Tasks
Video classification
Languages
Arabic, Romanian, English, German, Hindi, Spanish, Russian
Size
10K–100K rows
Downloads this month
43K
Likes
0

How to load it

from datasets import load_dataset

ds = load_dataset("Wan-fen52/MAVOS-DD")