Typographic self-portrait generation in an interactive museum installation

This article presents Photomaton, an interactive museum installation that generates typographic self-portraits of visitors using computational design methods. Operating in a photo booth configuration, the system captures a photograph of the user and renders a portrait composed exclusively of glyphs, combined with a user-selected literary text. We describe the design, development, and evaluation of Photomaton, including three generative approaches for typographic portrait creation based on text, words, and glyphs, as well as a template-based method for automatic typesetting and layout generation. A prototype was evaluated with 46 participants, demonstrating efficient portrait generation, high recognisability of user facial features and strong user satisfaction. The installation has been successfully deployed as part of the permanent Torre Literária exhibition, dedicated to Lusophone literature and inaugurated in Vila Nova de Famalicão, Portugal, in late 2020. The contributions of this work include the development of an interactive system for real-time typographic self-portrait generation; the design and evaluation of multiple generative approaches that explore the balance between textual readability and image recognisability; the integration of automatic typesetting and template-based layout generation within a fully functional interactive installation; and the demonstration of its applicability in real-world museum contexts, including support for print-on-demand customised artefacts for visitors. Additionally, this work provides a documented case study demonstrating the deployment of computational design methods in a public museum setting.