Image Tampering Datasets, Image Manipulations Datasets... A repo intend to collect most of datasets (training and evaluation) for the image forgery detection and localization.
图像篡改检测定位数据集收集
Paper: IMD2020: A Large-Scale Annotated Dataset Tailored for Detecting Manipulated Images
Project Page: http://staff.utia.cas.cz/novozada/db/
Download: http://staff.utia.cas.cz/novozada/db/
Appear in: [1]P. Zhuang, H. Li, S. Tan, B. Li, and J. Huang, “Image Tampering Localization Using a Dense Fully Convolutional Network,” Ieee T Inf Foren Sec, vol. 16, pp. 2986–2999, 2021, doi: 10.1109/tifs.2021.3070444.
-
Artificial PS dataset with post-processing on boundary (PS-boundary)
-
Artificial PS dataset with arbitrary post-processing (PS-arbitrary)
-
NIST 2016 dataset (NIST-2016)
Paper: The 'Dresden Image Database' for benchmarking digital image forensics SAC '10
Official: http://forensics.inf.tu-dresden.de/dresden_image_database/ (Cannot be ACCESSED)
DEFACTO database
- Dataset contains over 200,000 forged images with different kind of manipulations.
- Paper Link
- Dataset Download Link
https://www.ee.columbia.edu/ln/dvmm/newDownloads.htm
Columbia Uncompressed Image Splicing DetectionColumbia Uncompressed Image Splicing Detection: https://www.dropbox.com/sh/786qv3yhvc7s9ki/AACbEEzGPrD3_y38bpWHzgdqa?dl=0
Columbia Image Splicing Detection Evaluation Dataset: https://www.dropbox.com/s/bo10et4p1zg08aj/ImSpliceDataset.rar?dl=0
Description: This dataset contains 220 realistic forgeries created by hand in modern photo-editing software (GIMP and Affinity Photo) and covers various challenging tampering scenarios involving both object insertion and removal...
Citation: Pawel Korus and Jiwu Huang. Multi-scale analysis strategiesin prnu-based tampering localization. Trans. Info. For. Sec.,12(4):809–824, apr 2017.
Page: https://pkorus.pl/downloads/dataset-realistic-tampering
Download: realistic-tampering-dataset.zip via google drive (1.7 GB) you'll need request access
Small image category database: 512 x 512, 200 image sets, 40 images per transformation type, total number of images with post-processed images = 10400. Large image category database (please contact us for downloading): 3000 x 2000, 60 image sets, 10/20 images per transformation type, total number of images with postprocessed images = 3120.
-
Paper: CoMoFoD — New database for copy-move forgery detection
-
Project Page: https://www.vcl.fer.hr/comofod/download.html
-
Download:
- https://www.vcl.fer.hr/comofod/comofod_small.rar (Official, for the 200 image sets of Small image category database)
- Large image category database, I would recommend you contact [email protected], a quick reply in one hour.
- Information Forensics and Security
- Forensics Dataset
- Pattern Recognition Lab, FAU, https://lme.tf.fau.de/category/dataset/
| Name | Paper | Project | Download | Type | Remarks |
|---|---|---|---|---|---|
| Columbia | |||||
| CASIA | |||||
| MICC | |||||
| IMD | |||||
| CoMoFoD | |||||
| COVERAGE | |||||
| GRIP | |||||
| FAU | |||||