Automated Measures of Visual Diversity: A Computer-Vision Approach to UN Women’s Instagram Content
(A2026-133924)

AUTHORS

Carolina Nicolas, University of Santiago Chile (USACH); Michele Girotto, Universitat de Barcelona

ABSTRACT

This study examines how the visual representations shared on UN Women’s official Instagram account align with theoretical and empirical frameworks that guide diversity in advertising, offering conceptual, methodological and empirical insights into inclusive digital communication within international organisations. Adopting a quantitative, computational approach based on automated image analysis, the research evaluates gender, age, and skin-tone representation across the organisation's visual output. The results reveal an average global inclusivity index of 0.5092, indicating a moderately inclusive level of visual communication. The analysis highlights a central challenge for UN Women: reconciling its legitimate mandate to foreground women's visibility with the growing expectation to project intersectional, comprehensively inclusive narratives. By applying computational image analysis within an institutional context in which gender equity is the core mandate, the study makes an original contribution to research in social marketing and institutional communication. It illustrates the value of data-driven methods for representational coherence in digital environments.