Digital media objects play a central role in everybody's life. Anybody can capture images, videos and audio. They are used in everyday communications and they are widespread online. However, all digital media can be easily manipulated by anyone with easy-to-use software tools. As the diffusion of manipulated media can lead to serious consequences (e.g., fake news, warfare propaganda, etc.), the Multimedia Forensics (MMF) community keeps working to develop tools to assess the authenticity and integrity of all media.

Despite the huge effort put by researchers, the astonishing advances in the artificial intelligence field and the new trends in digital media creation are posing novel challenges. On one hand, adversarial machine learning has shown that it is possible to craft attacks to undermine the performance of any detector. On the other hand, media are evolving. The goal of FOSTERER (Robust Multimedia FOrensic Solutions To Enhance ModeRn MEdia TRustworthiness in Adversarial Conditions) is to develop new forensic tools that prove robust in a modern scenario in which forensic detectors work in a naturally adversarial environment.

To do so, FOSTERER will consider both targeted and non-targeted attacks. To solve the cat-and-mouse game in which new defenses trigger new attacks, it will develop MMF defenses based on a series of robust detectors. The development of robust MMF tools will have a positive impact on enhancing media trustworthiness, thus helping in preventing fake news spreading, false propaganda, and enabling objective moderation over social media.