Imagga’s AI will Organize and Moderate Large Media Archives to Improve the European Media Landscape

In times where data is everything, AI solutions can be leveraged by media companies to improve pretty much every part of the media value chain – from helping content creators be more creative to helping readers find the content they are looking for and using meta-tagging to monetize content. In the field of media, however, the European AI landscape is still very fragmented and there is a need for the establishment of some kind of networking infrastructure that will bridge the gaps between academia and industry. Imagga, the Bulgarian image recognition startup that provides API solutions and customizable ML technology for image tagging and categorization, is one of the 30 participants in a consortium that conducts such а high-impact project for using AI in media advancements. 

Imagga’s platform of cloud-based and on-premise APIs for automated image and video tagging has been used by more than 15k developers and 250 businesses worldwide. The company started its cloud API in 2011, and its flagship auto-tagging and auto-categorization technologies in 2013.

The AI4Media project is funded with a budget of €12m by the EU’s Horizon 2020 Research and Innovation Programme and its core mission will be to put the latest developments in AI in use for the European media industry and to ensure that ethical and trustworthy technology is being deployed in future AI advancements. As it involves leading players from the AI and media sectors, the project also envisions to become a Center of Excellence and develop a network of researchers across Europe. Besides Imagga, the consortium consists of 11 other industrial organizations among which IBM, Deutsche Welle (DW), and F6S, 9 universities, and 9 research centers.

The benefits of meta-tagging for media 

The Head of AI in Imagga, Georgi Kostadinov, explains that over the years many of the largest media companies in Europe and around the world have accumulated and continue to accumulate huge digital archives and collections of photos and videos. Because these collections are built gradually over months and years, often by different departments of media companies, they usually do not have good organization or have little or no metadata such as tags, categories, and other types of annotations. According to him, the organization of these media archives would provide media companies with additional opportunities for monetization and would allow them to reuse these archives and transform them into new and useful content.  

Kostadinov also outlines that both large traditional media companies and online media platforms enable the generation and sharing of content by end-users. Although this type of content helps to create new channels for monetizing the platforms, the lack of its active moderation hides many risks and threats, including fines and sanctions. “This type of hybrid archives needs a quality solution for automatic content moderation, often working in real-time, to be able to protect end-users and comply with laws and regulations, both in different countries and on the Internet,” explains Kostadinov. 

What will be the role of Imagga’s AI solutions?

During the last couple of years, Imagga has developed Photobox which is an easy-to-use tool for searching and discovering visual content that allows automatic assignment of images and videos to different categories that can be used for concept recognition in the media and entertainment industry. As part of the AI4Media project, Imagga will build on this type of solution to create another solution that will be curated to automatically organize photo and video content used by media companies.

The technologies that Imagga is developing within the project will be tested on target groups of real users from media companies, some of which part of the AI4Media consortium such as Deutsche Welle and RAI – Italy’s largest radio-television broadcasting company. “By developing a single and comprehensive solution for media companies that automatically organizes and moderates their media archives, we believe we will contribute to the creation of a better AI ecosystem that will have a positive impact on the media sector and will show the added value of applying this type of technology,” adds Georgi Kostadinov.