Skip to the content.

Synthetic and Manipulated Overhead Imagery For Forensics Research

Brandon B. May, Kirill Trapeznikov, Shengbang Fang, Matthew C. Stamm
Comprehensive Dataset of Synthetic and Manipulated Overhead Imagery For Development and Evaluation of Forensic Tools
https://arxiv.org/abs/2305.05784

@inproceedings{may2023comprehensive,
  title={Comprehensive Dataset of Synthetic and Manipulated Overhead Imagery for Development and Evaluation of Forensic Tools},
  author={May, Brandon B and Trapeznikov, Kirill and Fang, Shengbang and Stamm, Matthew},
  booktitle={Proceedings of the 2023 ACM Workshop on Information Hiding and Multimedia Security},
  pages={145--150},
  year={2023}
}

👉 ⬇️ Dataset Download

Dataset Overview

Sources

Pristine Tile Providers

City Conditioning

Inpainting Category

Inpainting Category Region Type Manipulated 🔄 Pristine Mask
Greenspace Bezier
Buildings Bezier
Greenspace GraphCut
Buildings GraphCut

Inpainting Size

Inpainting Area Size Manipulated 🔄 Pristine Mask
Exra Small
Small
Medium
Large

Fully Synthetic Categories

Dataset Stats

Train Set

The train set consists of 10,000 images in total:

Each image has a corresponding binary mask image that indicates the manipulated region with the color white.

Test Set

The test set consists of 3,150 images in total:

Each image has a corresponding binary mask image that indicates the manipulated region with the color white.

Dataset Download

⬇️ Dataset Download

Metadata CSV Format

Each dataset is provided with a metadata.csv file which contains detailed information about each image. The following columns are particularly relevant for training and testing:

Additional Generation Modes

Partial Manipulations

Natural Disaster Simulation

Contact

For more information: Kirill Trapeznikov, kirill.trapeznikov@str.us; Brandon May, brandonbmay@gmail.com

Acknowledgment

This material is based upon work supported by DARPA under Contract No. HR0011-20-C-0129. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of DARPA.

DISTRIBUTION A. Approved for public release: distribution unlimited.