NasrinImp/COCO_AI

视觉计数图灵测试（VCT²）数据集旨在对AI生成图像检测（AGID）在先进文本到图像模型时代的挑战进行基准测试和分析。该数据集提供了由当代扩散和基于变换器的模型生成的图像的大规模、多样化的集合，包括Stable Diffusion 2.1、Stable Diffusion XL、Stable Diffusion 3、Stable Diffusion 3.5、DALL·E 3和Midjourney v6生成的图像，以及来自COCO数据集的人类创作的参考图像。每个样本包括一个文字描述（caption）、一张真实的COCO图像和多个模型生成的图像。The Visual Counter Turing Test (VCT²) dataset aims to benchmark and analyze the challenges of AI-generated image detection (AGID) in the era of advanced text-to-image models. The dataset provides a large-scale, diverse collection of images generated by contemporary diffusion and transformer-based models, including Stable Diffusion 2.1, Stable Diffusion XL, Stable Diffusion 3, Stable Diffusion 3.5, DALL·E 3, and Midjourney v6, along with human-created reference images from COCO. Each sample includes a caption, a real COCO image, and images generated from multiple models.