We introduce a novel deep neural architecture for image copy-move forgery detection (CMFD), code-named BusterNet. Unlike previous efforts, BusterNet is a pure, end-to-end trainable, deep neural network solution. It features a two-branch architecture followed by a fusion module. The two branches localize potential manipulation regions via visual artifacts and copy-move regions via visual similarities, respectively. To the best of our knowledge, this is the first CMFD algorithm with discernibility to localize source/target regions.
In this repository, we release many paper related things, including
- a pretrained BusterNet model
- custom layers implemented in keras-tensorflow
- CASIA-CMFD, CoMoFoD-CMFD, and USCISI-CMFD dataset
- python notebook to reproduce paper results
The entire repo is organized as follows:
- Data - host all datasets
- *CASIA-CMFD
- *CoMoFoD-CMFD
- *USCISI-CMFD
- Model - host all model files
- ReadMe.md - this file
Due to the size limit, we can't host all dataset in repo. For those large ones, we host them externally. *indicated dataset requires to be downloaded seperately. Please refer to the document of each dataset for more detailed downloading instructions.
The original model was trained with
- keras.version = 2.0.7
- tensorflow.version = 1.1.0
we also test the repository with
- keras.version = 2.2.2
- tensorflow.version = 1.8.0
Though small differences may be found, results are in general consistent.
If you use the provided code or data in any publication, please kindly cite the following paper.
@inproceedings{wu2018eccv,
title={BusterNet: Detecting Image Copy-Move Forgery With Source/Target Localization},
author={Wu, Yue, and AbdAlmageed, Wael and Natarajan, Prem},
booktitle={European Conference on Computer Vision (ECCV)},
year={2018},
organization={Springer},
}
- Name: Yue Wu
- Email: yue_wu[at]isi.edu