Overcoming the challenges of AI in Media

A busy summer has flown by here at the DPP, so it’s hard to believe that it was just eight weeks ago we hosted the AI in Media event with our partners Digital Catapult, Microsoft, and Cognizant. But in that time I’ve found the subject of artificial intelligence and machine learning cropping up again and again, as we discuss topics such as media in the cloud, content preparation, and libraries and archives.

Over and over we hear that AI is beginning to revolutionise the generation of metadata and the processing of media. But we also hear about the same common problems, such as algorithms that aren’t sufficiently tuned for specific use-cases, or metadata overload. These are problems that our AI in Media solution providers are attempting to address, to deliver machine learning solutions that meet real business needs.

Delivering solutions

Those delivering solutions include companies like Newsbridge, which demonstrated interfaces designed to cut through duplication to show journalists the most relevant metadata results, or Data Language which enables content owners to deploy AI to classify and tag content using their own vocabularies and dictionaries. Companies like Wirewax which creates highly specialised tools to automate common processes like programme segmentation or asset deduplication. And those like Storyfit which use automated script analysis to predict content success, or Adlede, which uses content analysis to automatically place relevant, appropriate, and brand-safe advertising around programmes.

These companies were all lucky enough to pitch at our AI in Media event. With such a limited number of pitch spaces available, we had to be highly selective of the companies we showcased. In many cases, we chose to feature new or small companies who you’re less likely to know already, but other DPP members are developing great solutions too, including:

• Dalet, which integrates AI metadata generation into asset management workflows with Media Cortex
• Eluv.io, which is using ML to enable machines to understand video
• Limecraft, which can integrate and reconcile different sources of metadata to create a higher level meaningful description of content
• Prime Focus Technologies, which can solve problems including automatic generation of promos and artwork, and broadcast segmentation, through its Vision Cloud platform
• Sundog Media Toolkit, which uses AI to assist in colour grading and for enhancing security in the dubbing process
• x.news, which provides journalists with new tools for research

And of course, there was a whole lot more innovation on display at the inaugural DPP Innovation Week.

Working together

But the AI in Media event wasn’t just a one-off showcase. We heard from content companies and solution providers who are committed to developing practical solutions to real business problems. A number of them are talking to each other, forming partnerships, or scoping out proof-of-concept projects.

The DPP, along with Digital Catapult, Microsoft, and Cognizant, are here to help them, with the right business connections, project support, and technical backup. When companies with real business challenges and great content get together with companies with technical expertise and domain knowledge, exciting things can happen.