You are here
Research & Development
AFP Medialab is a small multi-disciplinary team aimed to boost innovation at AFP. It participates in several R&D projects with French or European funding. AFP Medialab is the main developer of the InVID-WeVerify browser extension, one of the main toolbox to debunk disinformation.
The Medialab is participating in the WeVerify project. This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement n°825297.
Online disinformation and fake media content have emerged as a serious threat to democracy, economy and society. Content verification is currently far from trivial, even tor experienced journalists, human rights activists or media literacy scholars. Moreover, recent advances in artifical intelligence (deep learning) have enabled the creation of intelligent bots and highly realistic synthetic multmedia content.
WeVerify aimed to address the complex content verification chalenges through a participatory verification approach, open source algorithm, low-overhead human-in-the loop machine learning and intuitive visualizations.
This plugin has been designed as a verification “Swiss army knife” helping journalists, fact-checkers, researchers and human rights defenders to save time and be more efficient in their fact-checking and debunking tasks on social networks especially when verifying videos and images or analysing disinformation propagation on social networks. This is a new version redesigned and enhanced with new tools, developed within the Horizon 2020 EU innovation action WeVerify.
AFP Medialab is the main developer and designer of the InVID-WeVerify verification plugin (launched and enhanced within the two eponym EU H2020 projects). This verification toolkit has attracted more than 57000 active weekly users worldwide (in 202 countries) at the end of November 2021. The plugin empowers journalists, factcheckers, teachers, academics, law and enforcement teams, and human rights defenders to debunk disinformation. The plugin has won in 2021 the first prize of the US Paris Tech award from US State of department GEC and Atlantic Council DFRLab.
The goal of the European YouVerify! project is to address the crucial challenges brought by disinformation and help young people develop both knowledge and know-how to make the difference daily between genuine and manipulated images and videos.
YouVerify! capitalizes on the results of previous Youcheck! pilote project (2019-21) by focusing on critical visual literacy training so as to scale up and disseminate competences in image and video verification across Europe. It uses the InVID-WeVerify plug-in, originally developed for journalists, to foster information resilience and debunking capacities among young people and youth educators (teachers, librarians, youth workers…).
The project is funded by the European Commission, within the Media Literacy for All Programme (2021). It is operated by the AFP in France, UNED in Spain, SNSPA in Romania, and Savoir*Devenir in France, which coordinates the project.
The goal of the EU-funded YouCheck! project is to raise awareness of the public at large and the education and media literacy community about disinformation and “fake news” as a threat to democracy.
YouCheck! aims more specifically to empower European citizens with the latest advances in image and video verification so that they can be actors of their own resilience strategies when confronted to information disorders.
YouCheck! answer such goals by adapting to media information literacy the InVID-WeVerify plugin already used by thousands of journalists and fact-checkers, especially for visual verification of images and videos.
InVID: web video verification
InVID developed a suite of tools to verify videos spreading on social networks, to analyse their reliability, their newsworthiness, and to check the reproduction rights with uploaders. In July 2017, AFP Medialab released publicly the InVID verification plugin as a free and open source toolbox to help fact-checkers verifying online videos.
ASRAEL: datification of events
The ASRAEL project develops a media events search engine that a journalist can query filtering on events attributes like asking a set of documents on earthquakes in Asia of magnitude > 7. This project is using Natural Language Processing (NLP) and semantic web techniques.
The internal projects
AFP Transcriber is an automatic speech analysis and audio transcription tool, allowing synchronization of text with video.
AFP 4W is a media event search engine and a generator of web pages, geographical mashups and interactive timelines.