Multi-Scale Tampering Maps Dataset

This dataset contains multi-scale tampering probability maps obtained by running a standard sliding-window PRNU detector (with central-pixel attribution) on a set of 136 tampered images. The images represent realistic forgeries, created by hand in modern photo-editing software (GIMP and Affinity Photo). The images were captured by four different cameras: Sony alpha57 (own dataset), Canon 60D (courtesy of dr Bin Li), Nikon D7000, Nikon D90 (RAISE dataset).

The dataset is intended for testing multi-scale fusion algorithms. Each forgery case has the following data:

  • seven 240 x 135 px multi-scale tampering probability maps (stored in .mat files)
  • 480 x 270 px ground truth tampering map (_mask.png files)
  • 480 x 270 px RGB thumbnail of the tampered image (.tif files)

A preview of an example forgery is shown below.

In order to minimize the size of the dataset, the tampering probability maps are stored with uint16 precision. Use Matlab's function im2double to obtain floating-point numbers in range [0,1]. The thumbnails can be used to guide the localization / fusion schemes. Full-size images are available as a separate dataset.

Using the Dataset

The dataset comes from our paper Multi-scale Analysis Strategies in PRNU-based Tampering Localization and can be used only for educational and research purposes. If you use it in your research, please cite the following paper:

@article{Korus2016TIFS,
  Author = {P. Korus and J. Huang},
  Journal = {IEEE Trans. on Information Forensics \& Security},
  Title = {Multi-scale Analysis Strategies in PRNU-based Tampering Localization},
  Year = {2017}
}

The dataset can be easily used in conjunction with our multi-scale analysis toolbox. You can obtain a copy as follows:

> git clone https://github.com/pkorus/multiscale-prnu

Once you have cloned the repository, you can easily download this dataset using:

> ./configure.py data:maps

Example use of these maps can be found in the demo_fusion script. A simple benchmark is also available (demo_benchmark script). For more information, please refer to the documentation of the toolbox and the above-mentioned paper.

Downloads