MediaVerse aims to enable all sorts of content creators, from traditional publishers and freelance creators or artists to anyone who wants to share their ideas, to create and share their media while keeping control of their intellectual property rights. It aims to provide a set of tools to support all steps in this process, namely co-creation tools, social analytics tools, a decentralised network to share the media, AI-supported tools for content analysis, blockchain-enabled tools, and automated language translation and other tools to facilitate the creation of accessible media.

Our team develops a number of technologies and services that facilitate multimedia content discovery, identification, and moderation. This involves the detection of near-duplicate multimedia, hate speech, nudity, misleading content, and content with violent and gruesome scenes.

Oct 2020 - Sep 2023

AI4Media aims to address the challenges, risks, and opportunities that the wide use of AI brings to media, society, and politics. The project aspires to become a centre of excellence and a wide network of researchers across Europe and beyond, with a focus on delivering the next generation of core AI advances to serve the key sector of Media. Its goal is to ensure that the European values of ethical and trustworthy AI are embedded in future AI deployments and to reimagine AI as a crucial beneficial, enabling technology in the service of Society and Media.

Our team develops a number of technologies and services that facilitate the detection of AI manipulated and synthetic content in multimedia. This involves deep fakes detection and image/video forensics.

Sep 2020 - Aug 2024

WeVerify aims to address the complex content verification challenges through a participatory verification approach, open source algorithms, low-overhead human-in-the-loop machine learning and intuitive visualizations. Social media and web content will be analysed and contextualised within the broader online ecosystem, in order to expose fabricated content, through cross-modal content verification, social network analysis, micro-targeted debunking and a blockchain-based public database of known fakes.

We develop technologies and services for the detection of deep fakes, the identification of multimedia reuse, the improved understanding of social media context around online images and videos, and the visualization of social media discussions on topics affected by disinformation.

Dec 2018 - Nov 2021 aims to disrupt the news media landscape by providing a set of tools that empower journalists in creating engaging news stories through targeted video mashups. The project will develop a service-based platform that will manage content from professional videos and User-Generated Content (UGC). The platform will allow the collection of videos related to a news event, and then their automatic annotation based on content and viewpoint, their organization, and mixing into summaries to be integrated in news stories. Beyond a simple “money saving through automation” platform, aims to leverage monetisation for tomorrow’s media by delivering more engaging news streams, supported by easy to consume video summaries presenting the key concepts and viewpoints for the unfolding stories.

Our team developed a number of technologies and services that facilitate the filtering and annotation of videos for the video summarization platform. These include near-duplicate video retrieval, video segmentation, logo detection, and context-based video stance detection.

Mar 2019 - Aug 2020

In this context, the main objective of the TENSOR project is to provide a powerful terrorism intelligence platform offering LEAs fast and reliable planning and prevention functionalities for the early detection of terrorist organised activities, radicalisation and recruitment. The platform integrates a set of automated and semi-automated tools for efficient and effective searching, crawling, monitoring and gathering online terrorist-generated content from the Surface and the Dark Web.

Our team developed the image forensics component of the TENSOR platform, with the goal of helping analysts detect and localize digital manipulations on images.

Sep 2016 - Aug 2019

The InVID innovation action develops a knowledge verification platform to detect emerging stories and assess the reliability of newsworthy video files and content spread via social media.

Our team developed a number of services for video verification, which was the core capability of the InVID platform. These include video forensics, near-duplicate video retrieval, logo detection and aggregation and analysis of video context.

Jan 2016 - Dec 2018

We develop tools and services that aid in Social Media verification. We look at verification from a journalistic and enterprise perspective.

We developed methods and tools that supported the multimedia verification needs of the REVEAL platform. These include a web multimedia crawler, a media clustering and summarization component, an image verification assistant based on image forensics, a tweet verification assistant based on semi-supervised learning, a Twitter account classification service based on graph embeddings, and a text-based location estimation service based on language models.

Nov 2013 - Dec 2016

SocialSensor offers technologies for analyzing the activity in social networks with the goal of discovering trending and high-quality content, identifying influencers, sentiment around discussed topics, and establishing links between different online sources. In addition, the project creates new means of retrieving, delivering and presenting media content and information in easily digestible ways and in mobile settings.

Our team prototyped a real-time framework for collecting, mining and searching streams of social multimedia content. This includes services for topic and event detection, community detection and analysis, large-scale multimedia indexing and search, and a web-based dashboard for social media analytics.

Oct 2011 - Oct 2014