Unpunished Expeditions

Marcel Top and Louis Delbarre as the co-authors of this project.

This project was made with Louis Delbarre.

Unpunished Expeditions wants to showcase the normalisation of extreme right symbolism in the online sphere and social media through the use of OSINT, data collection and image collages.

The project is an in-depth analysis of the recent banalization of the extreme right discourse, ideas, speech, and symbolism in the online sphere. This project aimed to recognize, categorize, and study this phenomenon, giving a wide variety of tools and resources to identify and fight the increasing presence of far-right groups online and in the streets. The recent presidential elections in France saw a radicalization of right-wing speeches and shifted the political landscape to the right. The small far-right groups became normalized very quickly and were offered access to very mainstream media.

The project is comprised of both gathered visuals and datasets, and is divided into four distinct stages;

Data and image collection:

The initial stage involves the collection of data and images, primarily sourced from social media platforms such as Facebook, Instagram, and Telegram channels. Our team infiltrated French far-right groups to obtain the images that were being shared both publicly and amongst members. The result of this stage was a substantial collection of images, featuring recurring subjects, symbols, and individuals. We categorized the images according to various themes based on the repetition of these elements.

Analysis of data:

We took all the data and images and started to analyse them. We took some of the images and put them into context, looking for the places where the images were originally taken. Using OSINT methods, we managed to geolocate most of the images downloaded. We ended up putting the social media posts into a bigger picture to give more context to the political situation and their presence all over France.

We categorized our research on the three most recurrent themes we found so far:

1. the three-finger salute

2. the Celtic Cross

3. the skull facemask

3D Model and AI:

As part of our effort to study and identify far-right symbols, we created 3D models of these symbols. By creating these models, we gained a deeper understanding of the construction of these symbols and how they may appear from different angles; lighting and perspectives. This process allowed us to recognize these symbols more easily in future imagery or still images, contributing to our overall ability to identify and track the presence of far-right messaging and propaganda. Through 3D models we are training an Object Detection Algorithm to recognise these symbols within imagery and videos shared online. The use of 3D models as a learning tool has been instrumental in our efforts to combat hate speech and extremist ideologies.

Extreme right imagery as a tool to combat extreme right symbolism in the online sphere:

By using, altering and exposing these far-right imagery, we hope to create, besides a multimedia project, a tool for people to understand and educate themselves about the normalisation of such symbols and the growing influence of these hate groups.

This project aims to tackle this growing issue by bringing it to light and developing a solution to combat it.