Â鶹Éç

As artificial intelligence is redefining how we see and understand the world, the Â鶹Éç Research & Development computer vision team is leading the charge. We’re applying cutting-edge AI tools to real-world challenges in broadcasting - from making virtual production affordable, to enriching decades of archive footage, to protecting the identities of those who share their most vulnerable stories.

At the Â鶹Éç, we’ve always believed in the power of storytelling - stories that inspire, inform, and connect people around the world. But as technology evolves, so too does the way we tell those stories. In the R&D computer vision team, we’re exploring the next frontier: how artificial intelligence can unlock new creative possibilities, make production more accessible, and uphold the Â鶹Éç’s values of trust and creativity.

Our mission is simple: to use AI responsibly and imaginatively to serve audiences better and support the people who make great programmes.

Transforming production with virtual creativity

Television production has traditionally been expensive, logistically complex, and resource-intensive. That’s why our team has been pioneering the application of novel view synthesis (NVS) for broadcast — a powerful AI technique that creates realistic 3D scenes from simple videos or photos. 

Imagine capturing a real-world location with just your phone and then turning it into a fully navigable virtual environment within hours. With NVS, small or remote productions can now achieve the kind of visual storytelling that once required large studios, green screens, and huge budgets.

By using open-source software and commodity hardware, we’re helping producers bring creative ideas to life faster, cheaper, and more sustainably. It’s not just about cost-saving, it’s about democratising production. As we prepare to showcase this work at , we’re collaborating with the to test our tracking and projection pipelines for future landmark series. That will allow us to take NVS technology out of the lab and into real-world production, helping us to refine our tools to better support producers in the field. Bridging the gap between research and production means teams can focus more on creativity and storytelling, and less on the time and cost of setup. It’s an exciting time for the team - not only innovating behind the camera but redefining what’s possible in front of it.

Bringing the archive to life with intelligent metadata

The Â鶹Éç’s Natural History Unit holds one of the world’s most extraordinary archives, decades of footage capturing the wonders of our planet. Yet, much of this treasure trove remains underutilised, simply because of the challenge of cataloguing and searching such vast amounts of content.

That’s where our metadata enrichment work comes in.

We’re developing AI-powered tools that can automatically generate descriptive metadata for wildlife footage; identifying species, detecting anomalies, and tagging content to make it searchable and usable for future storytelling. By making content more accessible, we’re enabling producers, researchers, and filmmakers to rediscover powerful moments and weave them into new narratives.

Our novel view synthesis and metadata enrichment work is part of the EMERALD project, a 30-month collaboration funded by the EU and UK Research and Innovation (UKRI), exploring energy-efficient AI for media. It reflects our commitment to innovation that’s not only intelligent but sustainable. We‘re using our skills and expertise to make a real creative difference – keeping the Â鶹Éç’s incredible legacy in natural history programming alive while finding new ways to tell its story.

Protecting identities with ethical AI

One of the most powerful applications of AI in media today isn’t about spectacle, it’s about safeguarding people. In current affairs and documentaries, contributors often share deeply personal, sometimes dangerous, truths.

Traditional methods of anonymisation such as blurring, pixelation, or shadowing can compromise both privacy and authenticity. They can even unintentionally suggest guilt or bias, undermining the integrity of a story. Our AI anonymisation project is pioneering a more human, ethical alternative.

Using advanced generative algorithms, we can now replace a contributor’s face with that of a consenting actor, while faithfully preserving every expression, emotion, and nuance of their interview. The result? A synthetic yet seamless version of reality that keeps contributors safe without diminishing the emotional truth of their story.

To ensure the highest standards of privacy, the system processes data and runs algorithms entirely offline, with all sensitive data securely erased after processing. By bringing these tools in-house, we’re not only enhancing trust and security but also delivering cost savings to documentary productions. Work like this is where the technical and ethical sides of AI meet, and we are ensuring progress goes hand in hand with care and responsibility.

Collaboration

Our work thrives on collaboration. We maintain close relationships with leading universities, including the , , and , as well as organisations such as , the , and the .

These relationships keep us at the cutting-edge of AI research, while grounding our work in real-world applications. They put our models through rigorous testing and peer review to ensure that they meet the Â鶹Éç’s standards of accuracy and integrity.

The road ahead

The AI Research team at Â鶹Éç R&D brings together a wide range of talents and perspectives to explore how creative thinking and advanced technology can shape the way millions of people experience stories every day.

Our work uses AI not to replace human creativity, but to amplify it. Whether you’re a researcher pushing the boundaries of generative models or a developer building scalable production tools, our work has visible, meaningful impact.

As AI continues to evolve, so will we. The next chapter of our work is about scaling our work across the Â鶹Éç and beyond. We aim to strengthen our collaborations with creative partners - ensuring every breakthrough is guided by transparency, ethics, and the Â鶹Éç’s public service values.

We’ll continue to share our findings, publish at leading conferences, and engage the public in conversations about AI transparency because trust in media is not just a Â鶹Éç priority, it’s something the world depends on.