2024 Accelerator Project

AI Media Production Lab 

In an IBC Accelerators first, the 2024 Programme featured three very special challenges under an umbrella theme of ‘AI Media Production Lab’, featuring an all-star cast of Champions and Participants exploring a series of specific AI concepts in various production scenarios within this exciting, fast-moving and disruptive area of media and technology. 

This encompassed researching and developing learnings around various AI tools and techniques to improve creativity and efficiency in production workflows, stamping out bias and AI hallucinations to embrace audience sentiments in content development, and exploring predictive video workflows in sports production, 

Supported by Champions: Zwart, EO (Evangelische Omroep), EBU, Verizon Business, ITV, Channel 4, YLE, BNNVARA, Rai, Paramount Global, Al Jazeera Media Networks, IET, World Freestyle Football Association, Vodafone Group

Supported by Participants: Pluxbox, Plan IX Labs, Respeecher, Xansr Software Private Limited, RKG Group, Somersault, Magnifi

The three Case Studies within IBC’s AI MEDIA PRODUCTION LAB include: 

Final Showcase Session

style="position: relative; display: block; max-width: 1220px;">

Generative AI in Action

​​​​The Challenge:

Are generative techniques ready to augment and accelerate how text and audio-visual materials are created for high-quality productions? And, how can media organisations, such as broadcasters or creative agencies, integrate Generative AI (GenAI) into their workflows to generate value? While numerous generative AI tools are available and rapidly evolving, many production companies lack the skilled teams and experience to effectively utilise them, especially when the need to combine multiple AI tools for high-quality output is essential. 

This creative IBC Accelerator, proposed by Champions RAI and EBU, and supported by ITV, IET, YLE, Verizon Business and Participants Pluxbox, Plan IX Labs, Respeecher, Somersault and Xansr Media, explored these questions to integrate a multitude of AI tools that were put through their paces towards creating an animated, proof-of-concept video from scratch. The project team also created an automated content creation workflow that enhanced real-time content customisation to streamline the process, thereby driving efficiency and innovation in the dynamic media broadcasting industry. 

Check out their Final Showcase session that successfully showcased highlights of their biggest learnings with what worked, and with the advances in graphics, image, video, voice and audio tools constantly evolving, where improvements are needed to produce the best possible broadcast-quality content.  

The POC Objectives:

  • Create and demonstrate a workflow using various Gen AI tools, from pre-production to production phases. 

  • Use this POC workflow to create a short clip, inspired by the characters and storyline of ‘Around the World in 80 Days’ by Jules Verne, set in a futuristic era. 

  • Document best practice learnings related to the usage and responsibility of AI tools from script writing to video, audio & sound generation.  

The POC Results:

  • The project team successfully created and demonstrated a workflow using a variation of Gen AI tools throughout the production pipeline. 

  • The POC resulted in an animated demo clip that showcased a futuristic take on ‘Around the World in 80 Days’ by Jules Verne. 

  • The POC also created an essential “Wall of Learnings” related to the utilization of generative AI tools for the different phases of production workflows (script writing, pre-production, assets production and integration).  

Final Proof of Concept Results

Read the 'Generative AI in Action' Case Study Deck Here.  

View Here

View All Additional 'Generative AI in Action' Documents Here.

Read More

AI Audience Validation Assistant (AAVA)

​​​​The Challenge:

This unique IBC Accelerator Challenge, proposed and led by leading Dutch broadcasters Zwart, EO and BNNVARA, with support from Champions IET, Al Jazeera and Channel 4 and Participant RKG Media, had a core focus aimed at the entire industry and audience perceptions to utilise AI as a tool to revolutionise the media landscape to combat bias, eliminate AI hallucinations to promote inclusion and diversity for audiences and focus groups at early content production and development stages, and ensure that content is truly inclusive and representative of diverse audiences.  

To explore these objectives, this multi-national consortium led intensive R&D to create a one-of-a-kind POC in ‘AAVA’, or, the ‘AI Audience Validation Assistant’. This full working demonstration POC experimented with AI personas and tested how a synthetic audience can deliver real value across media production as active contributors, foster a new era of content creation and consumption, and ultimately, tech for good. 

This POC also pushed the boundaries to understand and engage a target audience, paramount for media organizations, so they can produce content that resonates with all parts of that audience. Traditionally, success was only visible post-publication or through costly and time-consuming methods like focus groups. In an age where data is key, this project led in innovation by combining extensive audience data with advanced AI to blend and transform content creation, involving the audience in the development process. The project also explored solutions to turn passive viewers into active contributors through AI-driven personas that reflect society's complexity.  

This unique approach accelerated content development, infusing it with genuine societal insights. Essentially, the POC results showcased a vibrant, interactive platform where the public not only consumed media but also plays a role in its creation, and quest for diversity and inclusion, heralding a new era of public-engaged media.  

The POC Objectives:

  • Improve audience engagement: Develop a prototype to create relevant and resonant content that better aligns with audience preferences. 

  • Show the urgency of changing dynamics in the landscape and media consumption  

  • Foster innovation in media production: Leverage AI to pioneer new ways of creating audience-validated content, setting a new standard for engagement. 

  • Instant audience feedback: Provide real-time feedback during the content creation process to enhance relevancy and audience satisfaction. 

  • Increase inclusion and diversity: Ensure representation of diverse groups in media by creating AI-powered digital twins that reflect the complexity of society. 

The POC Results:

  • AAVA prototype demo, with positive user feedback: Users reported increased interest in adopting the AAVA prototype for future content creation, recognizing its potential in improving audience insights. 
  • Collaboration with industry experts: Partnered with media companies, broadcasters, and tech leaders to refine and validate the tool's functionality. 

  • Addressed ethical concerns: Started with developing measures to combat AI bias and ensure cultural sensitivity, with a focus on fair representation. 

  • Tested AI digital twins: Successfully generated and utilized AI digital twins to simulate diverse audience members, providing instant feedback on content ideas and scripts. 

Final Proof of Concept Results 

Read the 'AI Audience Validation Assistant (AAVA)' Case Study Deck Here. 

View Here

 

View All Additional 'AI Audience Validation Assistant (AAVA)' Documents Here.

Read More

Changing the Game: Predictive Generative AI

​​​​The Challenge:

For this IBC Accelerator Challenge, leading Champion Verizon Business Group conducted an innovative POC with leading global broadcasters Paramount Global, Vodafone Group, Al Jazeera, IET and the World Freestyle Football Association, and Participants Xansr Media and Magnifi, to explore how Generative AI (GenAI) could reshape live sports and event broadcasting by creating personalized content and enhanced viewer engagement. The consortium harnessed large language models (LLMs), curated datasets, metadata, and video analytics to deliver predictive video and audio content that aligned with each viewer’s preferences. 

A standout feature of the POC was a customised demo ‘the Sports Companion’ that facilitated emotionally intelligent, voice-based conversations, allowing users to interact with sports content in a way that resembled human dialogue. This added a layer of realism and personalization that enabled fans to engage with an actual football match, through AI-driven conversations that felt intuitive and natural. 

In addition to enhancing interactivity, the POC introduced a powerful accessibility feature. Through advanced live video processing, it was able to generate real-time audio descriptions and commentary for videos that previously didn’t have any audio, so live broadcasts can be more inclusive for viewers with visual impairments. 

By integrating thought leadership from the multi-national Champions and Participants along with using cutting-edge technologies like Azure AI and Azure OpenAI, this Accelerator demonstrated how the Media & Entertainment industry could revolutionize live-to-air storytelling and fan engagement. And while live at the IBC Accelerator Zone at IBC itself, the POC was delivered over Verizon’s Private 5G networks and leveraged Open Cache technology, which enabled the PoC to successfully showcase the potential for AI to create a more immersive, personalized, and accessible live event experience. 

The POC Objectives:

This proof-of-concept aimed to redefine how fans engage with live sports and events through advanced Generative AI. These objectives, combined to deliver a forward-thinking vision for the future of live sports and event broadcasting, used AI to create more immersive, interactive, and accessible fan experiences. The key objectives included: 

  • Personalized Content Creation: AI-driven voice, text, and video experiences were developed to tailor content to each fan’s preferences, delivering personalized and relevant engagement. 

  • Real-time Interactivity: Fans engaged in live, real-time interactions during events, influencing the content they experienced, from video highlights to live commentary, through conversational AI. 

  • Authentic Conversations: Emotionally intelligent, voice-based interactions were utilized to create natural, human-like dialogue that directly shaped each fan's video and audio experience. 

  • Enhanced Accessibility: AI-powered accessibility features, such as real-time audio descriptions for video content without existing audio, were incorporated to ensure inclusivity for viewers with visual impairments. 

  • Ethical AI Integration: Ethical considerations were thoughtfully integrated, ensuring AI was used responsibly with personal information and content subject matter, maintaining privacy and trust while delivering a personalized experience. 

The POC Results:

  • A workflow that will give Broadcasters, Brands and Sports franchises a better understanding of how to build highly accurate, enterprise- grade solutions for real-time fan engagement. 

  • Video based sports metadata generated for established sports like Football, via SuperSport Schools, which enabled exploration to look ahead to improve newer sports via the World Freestyle Football Association. 

  • An emotionally intelligent sports companion that fans can talk to, and which auto-created a persona based on the conversation, that stayed transparent to show the information it used to build the persona. All the information was in enterprise grade security with no data used for training. 

  • The POC also featured a personalised highlight reel, and a GenAI person audio description and commentary for the personalized highlights. 

Final Proof of Concept Results

Read the 'Changing the Game: Predictive Generative AI' Case Study Deck Here. 

View Here

View All Additional 'Changing the Game: Predictive Generative AI' Documents Here.

Read More

Supported by:

logo

Champions:

Participants: