furonghuang-lab/ETI_Competition_Data

ETI竞赛数据集：从NeurIPS 2024竞赛“擦除隐形”中收集的完整数据。该数据集旨在系统地评估当前最先进的无形图像水印技术对抗全球范围内的参赛者开发的多种去除攻击的鲁棒性。数据集包括原始水印图像、参与者提交的评估结果、攻击后的图像文件等，为数字水印、对抗性机器学习、图像取证和内容真实性研究提供了大规模的基准。

ETI Competition Data: The complete dataset collected from the NeurIPS 2024 Competition Erasing the Invisible. This dataset is designed to systematically evaluate the resilience of state-of-the-art invisible image watermarking techniques against a wide array of removal attacks developed by participants worldwide. It includes original watermarked images, participant submission evaluations, and attacked image files, providing a large-scale benchmark for research in digital watermarking, adversarial machine learning, image forensics, and content authenticity.