The
AutoSplice dataset is a text-prompt manipulated image dataset for media forensics. It leverages the large-scale language-image model to automatically generate and splice masked regions guided by a text prompt. The dataset contains 5,894 manipulated and authentic images. More details can be found in the paper (
link).
In order to access and utilize the AutoSplice dataset, you are required to consent to the terms outlined below:
(1) The AutoSplice dataset is for non-commercial research purposes only.
(2) You and your affiliated institution must agree not to reproduce, duplicate, copy, sell, trade, resell or exploit any portion of the images or any derived data from the dataset for any purpose.
(3) You and your affiliated institution must agree not to further copy, publish, or distribute any portion of the AutoSplice dataset or any derived data from the dataset for any purpose.
(4) You and your affiliated institution take full responsibilities of any consequence as a result of using the AutoSplice dataset, and shall defend and indemnify the authors or the authors’ affiliated institutions against any and all claims arising from such uses.
(5) The use of AutoSplice dataset in publications must cite the citation given below.
@inproceedings{jia2023autosplice,
title={AutoSplice: A Text-prompt Manipulated Image Dataset for Media Forensics},
author={Jia, Shan and Huang, Mingzhen and Zhou, Zhou and Ju, Yan and Cai, Jialing and Lyu, Siwei},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={893--903},
year={2023}
}
(6) The authors reserve the right to terminate your access to the AutoSplice dataset at any time.
Please put your ACADEMIC email address below (e.g., @buffalo.edu) and the download link will be sent to you once the form is accepted. If you have any questions, please send email to autosplice.dataset@gmail.com.
