Enhancing visual on-screen depiction through the application of computer vision technologies
In a groundbreaking initiative, the Nesta-Learning on Screen research series is using computer vision technology to measure on-screen representation and diversity in the UK screen industry. This innovative approach automates the labor-intensive task of representation measurement, providing reliable, scalable data to inform policy, production diversity goals, and social impact evaluation within the UK screen sector.
The method involves applying face detection and recognition models to identify individuals or demographic groups on screen, using image analysis to classify attributes related to ethnicity, gender, or age. By aggregating data over multiple shows, genres, or broadcasters, the series assesses overall representation, moving beyond manual coding.
This application of computer vision is significant as it enables systematic, large-scale, and objective measurement of representation, revealing statistical patterns and trends about who appears in UK screen media. The data gathered supports evidence-based discussions about diversity and inclusion in the industry.
The Nesta-Learning on Screen research series is an output of a pilot project conducted by Nesta in partnership with Learning on Screen. The survey was commissioned by the Creative Industries Council. While the specific documentation details are not included in the search results, the general method and intent of using AI-based computer vision for diversity measurement in screen content align with known cutting-edge research and innovation trends in media analytics.
It is worth noting that the report is not related to the use of computer vision for measuring on-screen representation in the UK screen industry, nor is it an output of Raphael and Bartolomeo's presentation at the International Federation of Television Archives (FIAT/IFTA) Conference 2020. The report does not mention any related research reports or draw inspiration from PEC research on generating creative diversity measures using machine learning, such as the project "She said more".
The UK screen industry has a diversity problem that persists despite progress. The diversity issue in the UK screen industry is rooted in structural inequalities. As such, diversity evaluation should consider not only on-screen presence, but also prominence and portrayal.
The report was provided by Creative PEC, an organization that offers independent research and policy recommendations for the UK's creative industries. The project team includes Raphael Leung, Bartolomeo Meletti, Dr Cath Sleeman, Gabriel A. Hernández, and Gil Toffell.
[1] The increasing deployment of computer vision in similar contexts to map and analyze representation at scale across broadcast media, such as enhanced multimodal understanding in GPT-5 and other AI models, supports the technical feasibility of this approach.
[1] The Nesta-Learning on Screen research series, employing artificial-intelligence-based computer vision, is addressing the diversity issue in UK screen industries by measuring and analyzing representation at scale, correlating with recent advancements in multimodal understanding with AI models like GPT-5.
[2] As technology advances and skills in AI and data analysis continue to grow, these innovative research series contribute to evidence-based education policies aimed at fostering creativity, inclusion, and talent development within creative industries, thus shaping a more equitable representation of society.
[3] Through the systematic measurement of on-screen representation, the series aims to provide extractable data, leading to benchmarks and annual reports that can influence policy, inspire industries to set ambitious diversity goals, and spur research on portrayal within various demographic groups.
[4] Recognizing artificial-intelligence as a key player in the future of industries, the Nesta-Learning on Screen research series highlights the potential for AI to transform research and education by enabling meaningful conversations around diversity, equity, and inclusion in creative industries.
[5] While the initial focus of the report was on screen content representation, it is crucial for future research to delve into the talent pipeline and explore the impact of AI-driven technology tools on creation and development in different sectors of the creative industries, providing valuable insights into supporting diversity and fostering a truly inclusive UK screen industry.
