MONOT is specifically designed for the Morphing Attack Detection (MAD) task.
It consists of a collection of morphed images, created through six different morphing algorithms. Morphed images exhibit high visual quality in terms of ISO/ICAO compliance, reproducing the operational scenario in which they are used as document photos. In addition, for each subject, several ”in the wild images” are provided resembling, for instance, the live acquisition at the airport gates.
MONOT dataset aims to provide a comprehensive and privacy-friendly resource for training and evaluating MAD systems.
In the wild images
For the generation of synthetic gate images, we focus on the Arc2Face model, an identity-conditioned face foundation model that can generate diverse photo realistic images from the ArcFace embedding of an individual.
Our primary objective is to generate new realistic images of an individual to simulate images captured live in an unconstrained scenario such as an ABC gate, where the head pose can be non-frontal, skin reflections are visible etc.
Download
To download the dataset, click on the following button.
To download the bona fide images, please download the ONOT dataset.
License and Reference
The MONOT dataset is released under the CC BY-NC 4.0 license: you may not use the material for commercial purposes and you must give appropriate credit (see the reference below). You are free to download and modify the dataset.
If you use this dataset please cite the following paper:
Borghi G., Di Domenico N., Ferrara M., Franco A., Latif U., Maltoni D. MONOT: High-Quality Privacy-compliant
Morphed Synthetic Images for Everyone. Proceedings of the 16th IEEE International Workshop on Information Forensics and Security (WIFS), 2024