ts0pwo/20K_real_and_deepfake_images_NA

这个数据集包含了用于评估deepfake检测框架的测试图像。原始数据集包含20,000张真实和deepfake图像，但由于2600张图像受到英国皇冠版权保护且未经授权不能复制，这些图像已被移除。该框架使用了4种机器学习模型，输入包括原始图像、错误水平分析（ELA）图像、噪声分析（NA）图像和主成分分析（PCA）图像。这些模型使用Tensorflow 2.26.2版本创建。在当前仓库中，存储的是噪声分析（NA）图像。

This dataset contains test images used to evaluate our deepfake detection framework. It originally included 20,000 real and deepfake images, but 2,600 files protected by the UK Crown have been removed due to lack of permission to reproduce them. The framework consists of 4 machine learning models, which take in original images, error-level analysis (ELA) images, noise analysis (NA) images, and Principal Component Analysis (PCA) images as input. These models were created using Tensorflow version 2.26.2. In this repository, the NA images are stored.