Exhibit Three: Demonstrative Evaluation of TV Character Influence

Exhibit Three: Demonstrative Evaluation of TV Character Influence

In the ever-evolving world of media, the role of technology is increasingly becoming crucial. A new approach is being explored to widen the evidence base of on-screen representation using computer vision. This innovative technique can automatically analyze visual media, such as videos, images, and UI screens, to detect and quantify diverse demographic features and content elements.

The concept was first illustrated using an episode of the popular British comedy show, Mock the Week (Season 15 episode 6), which was shown on BBC2 in March 2017 (originally aired in July 2016) and was produced by Angst Productions and Ewan Phillips. The short video clip was broken down into smaller components called 'scenes', and a face detector, a type of machine learning model, was applied to identify faces on screen and generate 'face tracks'. The time spent on screen, as a clear and big enough face, indicated the relative prominence of each character.

The second example used an episode from the American sitcom Black-ish Season 1 (ABC) to test the feasibility of generating character prominence metrics when there is a greater variety of camera angles and face sizes. This demonstration underscored the potential of computer vision to provide scalable, objective, and data-rich insights into representation patterns and content diversity.

Researchers, such as Raphael Leung, Data Science Fellow at Nesta, and Bartolomeo Meletti, Creative Director for CREATe at the University of Glasgow, are at the forefront of this exciting development. They believe that the methods can be extended to widen the evidence base around on-screen representation for four groups: diversity leads and monitors, content producers, editors and commissioners, and researchers.

The applications of this technology are far-reaching. Editors and commissioners can use computer vision to analyse character prominence before a show airs, allowing them to make informed decisions about representation. Content producers may represent new opportunities to create new features for viewers to look for major and minor characters.

Moreover, diversity leads and monitors can generate more frequent and richer data about representation, while researchers can analyse large datasets objectively without reliance on manual annotation or self-reported metadata, increasing transparency in diversity evaluation.

This innovative use of computer vision can provide empirical evidence of on-screen representation inequities or progress, informing diversity and inclusion initiatives. It can also assist content producers and commissioners in evaluating whether their work meets diversity goals by quantifying representation markers automatically.

As the application of computer vision to diversity evidencing is still emerging, advances in object recognition, keypoint detection, and vision-language models are rapidly expanding its capabilities. The integration of these technologies offers a practical, scalable approach to quantifying and understanding on-screen representation in ways that traditional manual methods cannot achieve comprehensively.

In related news, several research reports are available on International, Trade, and Immigration topics. For instance, a report details the migrant and skills needs of creative businesses in the UK, commissioned by the Creative Industries Council. Another report, titled "The impact of overseas mergers and acquisitions on UK video games industry", is part of the BFI’s research. The UK's departure from the EU has changed the way that British firms trade and work with European counterparts, as detailed in a report on post-Brexit migration and accessing foreign talent in the Creative Industries.

In the upcoming issue of ViewFinder - Learning on Screen's specialist online magazine dedicated to the moving image and education - will be exploring AI and its relationship with audiovisual media. The next frontier in the field of computer vision and media analysis is certainly promising.

1. In the media industry, researchers such as Raphael Leung and Bartolomeo Meletti are using computer vision to analyze on-screen representation in shows like Mock the Week and Black-ish.
2. This innovative technology can automatically detect faces on screen and quantify character prominence, offering scalable, objective, and data-rich insights into representation patterns and content diversity.
3. Editors and commissioners can use this technology to make informed decisions about representation before a show airs, while content producers may find new opportunities to create features for viewers.
4. Diversity leads and monitors can use computer vision to generate more frequent and richer data about representation, increasing transparency in diversity evaluation.
5. As the application of computer vision to diversity evidencing is still emerging, advances in object recognition, keypoint detection, and vision-language models are rapidly expanding its capabilities.
6. The integration of these technologies offers a practical, scalable approach to quantifying and understanding on-screen representation in ways that traditional manual methods cannot achieve comprehensively.
7. The upcoming issue of ViewFinder magazine will explore AI and its relationship with audiovisual media, while research reports are available on topics such as the migrant and skills needs of creative businesses in the UK and the impact of overseas mergers and acquisitions on the UK video games industry.

Read also: