This autumn, we organised a virtual competition "How to detect malicious use of video/photographic content online?". This competition provided an opportunity to meet talented young teams, who presented innovative concepts and prototypes aimed at countering malicious online activities, with a focus on altered images and/or video content. 

On December 10, the award ceremony took place, and the winning team from the University of Aveiro (Portugal) received a prize of 5 000 euros, provided by our main partner - Latvijas Mobilais Telefons

The winning project leverages data mining algorithms and machine learning in order to enhance the detection of video/photographic content used as part of online disinformation campaigns.
We also received very valuable contributions from two other teams. A team from the University of Turin proposed a dashboard that enables users to map disinformation networks. The same application makes it possible to find instances of malicious use of photographic content with a tailored image search feature. The other team represented the project, which is already operational and is helping journalists in the detection of disinformation online. Their entry consisted of a machine-learning approach to recognize extremist videos.
The selection committee was comprised of representatives from the NATO StratCom COELatvijas Mobilais TelefonsUS Embassy in Latvia, and Accenture.

We wish all three teams good luck in developing their prototypes further, and we hope this event was a useful opportunity to advertise their efforts. 

More pictures from the event are available here

Get in touch with the winning team to learn more about their solution: Dr António J. R. Neves, Assistant Professor, University of Aveiro, Portugal, email:

Or read about the team and their success on Portuguese media:
‚ÄčUniversidade de Aveiro ganha concurso da NATO para detetar mensagens extremistas (
NATO escolhe Universidade de Aveiro para combater extremistas (
Universidade de Aveiro ganha concurso para detectar mensagens extremistas (