DVijayan/vilhyra-vision-dataset

VISION — mirror

This repository redistributes the VISION dataset. It is a mirror for research and engineering convenience. All credit belongs to the original authors; nothing here is original work by the redistributor.

Attribution (required by the licence)

Shullani, D., Fontani, M., Iuliani, M., Al Shaya, O., & Piva, A. (2017). VISION: a video and image dataset for source identification. EURASIP Journal on Information Security, 2017(1), 15. https://doi.org/10.1186/s13635-017-0067-2

Original source: https://lesc.dinfo.unifi.it/VISION/ Produced by the Image and Communication Laboratory (LESC), Department of Information Engineering, University of Florence.

Licence

Creative Commons Attribution-ShareAlike 4.0 International (CC BY-SA 4.0) https://creativecommons.org/licenses/by-sa/4.0/

Verified against the upstream project page and the EURASIP paper on 2026-08-21.

What that means for anyone using this mirror:

• Attribution — you must credit Shullani et al. as above, and indicate if you made changes.
• ShareAlike — if you distribute an adapted version of this material, that adaptation must also be CC BY-SA 4.0. ShareAlike triggers on distribution, not on internal use: training on it privately does not oblige you to publish anything.
• No additional restrictions — you may not apply legal or technological measures that restrict others from doing anything the licence permits.

Whether trained model weights constitute Adapted Material under CC BY-SA is legally untested. The mainstream reading is that they do not. This mirror takes no position and offers no legal advice; if it matters to your use, take advice.

Contents

Camera-native images and videos from 35 devices, including native captures and platform-recompressed variants (WhatsApp, YouTube, Facebook). Its value is precisely that it is native-capture material: real EXIF, real quantisation tables, real device diversity — the properties that publisher-resized corpora such as Open Images do not have.

Why this mirror exists

It is the out-of-domain evaluation set for a forensic detection project. Downloading from the upstream host is slow from some regions; this mirror exists so evaluation runs are reproducible. If the upstream terms change, this mirror should be removed — please open a discussion if you are one of the authors and would prefer it taken down.

Corrections

• 2026-08-21 — README, licence declaration and attribution added. This repository was previously public with no licence, no attribution and no README, which did not satisfy CC BY-SA 4.0's conditions for redistribution. That was an oversight by the redistributor, corrected on discovery.