“AI is not just about cataloging and preservation, it’s about connecting your content to opportunity.” Jean Gabriel Minel, Head of Research and Development for the French National Audiovisual Institute (INA)
This quote sums up particularly well the key theme at the FIAT-IFTA World Conference held last month in Venice, Italy, and hosted by the RAI at the Palazzo Labia, a 17th century palace. Renowned for its exchange of ideas and best practices in archival science, the international symposium provided attendees the opportunity to discover innovative, AI-powered content management and archiving use cases and learn more about the latest technology trends from leading broadcasters, media organizations, and technology companies.
As a Gold Sponsor participant of the FIAT-IFTA World Conference, I was honoured to attend and represent Dalet. I had a chance to speak with attendees and show them how advanced content management, workflow orchestration, and best of breed AI services can better preserve, enrich, share, and monetize their archival content. With the combination of virtually limitless cloud storage and computing power, a new era is upon us.
Many of the attendees I met were archivists who hold degrees in library studies and have advanced knowledge of content lifecycle and asset management systems. During the conference, these tech-savvy content gatekeepers presented on how they are pioneering the use of AI to further enrich archived content in order to increase accessibility and generate new business opportunities. Many demonstrated that with automatic content discovery capabilities, media organizations can better monetize archival content and leverage their assets to gain a competitive edge over new entrants on their markets.
Applying AI to Media Workflows – The Need for a Plan
I found that most of those presenting about AI at the conference were trying to solve a similar issue: surface content easily to users with limited human intervention in logging the content. I can relate to this, as over the past 12 years working for Dalet, I have taught many of our customers about the importance of properly tagging content. With our latest generation toolset I now educate users on how to discover content that they need for their creations; should it be for creating promos, documentaries, news packages or highlights for sports.
I was honoured to present “The Dalet Living Archive: Utilizing AI in Archive Workflows” and demonstrate how Dalet solutions help media organizations streamline their workflows, reduce costs, and empower content creators with new tools to research, craft, deliver and promote their quality content to audiences on all platforms for consumption.
In the presentation, I demonstrated how Dalet Media Cortex automatically enriches content with AI services for:
- facial recognition
- speech to text
- entity detection
- content moderation
- object recognition
- OCR detection.
I showed that enrichment of the content triggers Dalet workflow orchestration to auto-create promos or highlights, and turn the AI-generated speech-to-text data into subtitles. I also showed that while a journalist types a story, Dalet Content Discovery detects entities such as people, places, and organizations as well as concepts, and then recommends content from the library with similar tags.
A Winning Example – Bundesliga Fußball Archives
Sportcast, a subsidiary of DFL (Deutsche Fußball Liga), which oversees the largest digital football archive in the world, manages the entire value chain from stadium to the world with the Dalet Galaxy Media Asset Management (MAM) and Orchestration platform. Sportcast has built a Media Portal on top of their Dalet system to connect and monetize all of the Bundesliga assets, leveraging the latest Dalet Orchestration technology to automate media processes and deliver more efficiently their partners.
With a grueling production schedule of 617 games each year, Sportcast is piloting AI-based indexing to optimize logging workflows and further enrich the massive amount of incoming content.
With great powers come great opportunities…
In addition to the Sportcast case study, there were other very interesting use cases presented at the conference, including the development of Visual Searches at the RAI, Ontology detection at RTVE, and facial/organization detection at RTS. The BBC demonstrated a new system that discovers content based on historical, cultural and other topical references linked to that particular day and publishes it to social media. They demonstrated how the system retrieved a funny clip from the 1950s for International Egg day. The asset would have never been found with a conventional search. The BBC’s AI system found it and within minutes automatically published it to Twitter. At the FIAT-IFTA World Conference, just a few months after they posted the Egg video, it had received 15 million views, further demonstrating how you can give life to old content.
With AI giving us so much data about an asset – thousands of tags for less than a minute of video for instance – a concern was brought up at the conference: would it become too difficult to choose content that is actually relevant?
The short answer, with the development of new tools utilizing faceted search capabilities, content discovery, workflow orchestration and auto-creation of content, the user decides what is important, giving viewers so many possible angles to the same story.
Prestigious awards for innovative projects
Held at the Scuola Grande San Giovanni, the FIAT-IFTA Gala Awards Ceremony recognized a number of Dalet customers including Danish Radio, SBS and RAI. Our congratulations to them and their teams on the accolades, and sincere gratitude for contributing to the discussions and innovation in the area of archives! See all Award Winners here!
We also would like to thank FIAT-IFTA and the RAI for organizing such an amazing event and look forward to next year’s conference in Croatia.
If you would like to learn more about our initiatives in building next-gen archiving workflows, please visit our dedicated solution webpage.
Jonathan Coutts-Zawadzki, Dalet's Head Of Knowledge Management, joined Dalet Technologies in 2006 in the development office as a QA Specialist, documentation expert and trainer. After moving to Dalet France, he became the lead trainer for many projects such as BBC Arabic and NBC Washington, to name a few. He then joined the Dalet Washington office in 2010 as the Project Manager of Dalet’s largest multilingual site, Voice of America. In 2012, Jonathan led Dalet Sales in South Asia as Country Manager of India, based in Delhi. After establishing the Dalet India office, Jonathan has recently taken on the role of Head Of Knowledge Management, working closely with the Sales and Pre-Sales teams to establish a good understanding of customer's needs and workflows.More Articles By Jonathan