Online images nowadays might be shared multiple times over Social Network platforms. To facilitate research on tracking social network origin of images, we collect two large-scale datasets: i) R-SMUD (RAISE Social Multiple Up-Download), and ii) V-SMUD (VISION Social Multiple Up-Download).

All images are shared at maximum three times through three platforms: Facebook (FB), Flickr (FL), Twitter (TW). The folders are named following this convention:

[P1]: original images are uploaded to platform [P1] and downloaded.

[P1]-[P2]: images in [P1] are uploaded to platform [P2] and downloaded.

[P1]-[P2]-[P3]: images in [P1]-[P2] are uploaded to platform [P3] and downloaded.

1) R-SMUD (download link).

50 RAW images are extracted from RAISE dataset [1] which can be downloaded from here. Those images are undergone top-left cropping to create 3 different sizes 377x600, 1012x1800, and 1687x3000 respecting 9:16 aspect ratio. Naming convention of cropped images is: original-[id]-[h]x[w].jpg, where id is the sequence number, h and w are height and width, respectively.

All cropped images are compressed using The Independent JPEG Group's JPEG software under six quality factors {50,60,70,80,90,100}. Compressed images are stored as: original/QF-[qf]/original-[id]-[h]x[w].jpg, where qf is the corresponding quality factor.

1) V-SMUD (download link).

510 JPEG images are extracted from VISION dataset [2] which can be downloaded from here. For each of 34 cameras, we select 15 images. Camera D12_Sony_XperiaZ1Compact is excluded since all of its images exceed 5 MBs which is the upper limit size allowed by Twitter. Images remain the same as they are in VISION. The list of used images is stored in original/original-jpeg.txt.
If you use one of these datasets, please cite our work as:

    title={Tracking Multiple Image Sharing On Social Networks},
    author={{Q.-T. Phan, G. Boato, R. Caldelli, I. Amerini}},
    booktitle={IEEE International Conference on Acoustics, Speech, and Signal Processing},


[1] D.-T. Dang-Nguyen, C. Pasquini, V. Conotter, G. Boato, RAISE - A Raw Images Dataset for Digital Image Forensics, ACM Multimedia Systems, Mar. 2015.

[2] D. Shullani, M. Fontani, M. Iuliani, O. A. Shaya, A. Piva, VISION: a video and image dataset for source identification, EURASIP Journal on Information Security, Dec. 2017.

All rights reserved. Multimedia Signal Processing and Understanding Lab. - University of Trento