Deepart Detection Database (DDDB)
收藏arXiv2023-02-28 更新2024-08-06 收录
下载链接：
http://arxiv.org/abs/2302.14475v1
下载链接
链接失效反馈官方服务：
资源简介：
Deepart Detection Database (DDDB) 是由西安交通大学、新加坡管理大学等机构联合创建的首个深度艺术检测数据库。该数据集包含高质量的传统艺术图像和由五个最先进的深度伪造模型生成的艺术图像，旨在探索一次性的深度艺术检测和持续的深度艺术检测问题。数据集中的传统艺术图像来自LAION-5B，而深度伪造艺术图像则由Stable Diffusion、DALL-E 2等模型生成。DDDB的应用领域主要集中在深度艺术的检测和版权识别，以解决深度艺术作品可能引发的伦理和版权问题。
The Deepart Detection Database (DDDB) is the first dedicated deep art detection database jointly established by institutions including Xi'an Jiaotong University and Singapore Management University. This dataset comprises high-quality traditional art images and art images generated by five state-of-the-art deepfake models, aiming to investigate the tasks of one-shot deep art detection and continual deep art detection. The traditional art images in DDDB are sourced from LAION-5B, while the deepfake art images are generated by models such as Stable Diffusion and DALL-E 2. The primary application scenarios of DDDB focus on deep art detection and copyright recognition, addressing the ethical and copyright issues caused by deep-generated artworks.
提供机构：
西安交通大学, 新加坡管理大学, 南安普顿大学, 哈尔滨工业大学, 鹏城实验室
创建时间：
2023-02-28
