Dataset · Yunncheng
AI-generated image detection test benchmark
Yunncheng/Mirage-Test
A collection of real and photorealistic AI-generated images across five domains for deepfake detector evaluation.
What's inside
Contains between 10k and 100k examples featuring images and English text labels (0_real and 1_fake). The generated pictures were created by modern models focusing on visual realism and span five distinct thematic domains. The dataset is intended exclusively for model evaluation and lacks a training split.
What you can do with it
- Benchmark classifiers for distinguishing real photographs from AI-generated images
- Test the robustness of synthetic content detectors across different thematic domains
- Tasks
- Image classification
- Languages
- English
- Size
- 10K–100K rows
- Downloads this month
- 28.7K
- Likes
- 3
How to load it
from datasets import load_dataset
ds = load_dataset("Yunncheng/Mirage-Test")