ki.portrait Gallery

Insight

The ki.portrait Gallery showcases a series of AI-generated portraits produced using DALL-E, presented at CityLAB’s summer fest event. Attendees had the chance to capture their portrait and leverage AI to generate distinctive artificial portraits by selecting parameters such as art style and materials. In a week-long experiment, I developed an interactive archive to document the project’s progress and display its outcomes. The archive, crafted as a single-page application, was built using Astro, TypeScript (vanilla), and WebGL.

Process

The goal was to craft an engaging archive showcasing the portraits and illustrating how various prompt parameters influenced the results. Opting for a 3D space, I utilized proximity to highlight similarities among portraits with shared user-selected parameters. When assigning high importance to factors like style or materials, distinct groups emerge, uniting portraits with similar artistic styles and material choices.

Personal role

While I was away during the summer fest and not directly involved in creating the photo booth or the software interfacing with DALL-E, I enthusiastically embraced the opportunity to contribute upon learning about the team’s motivation to publicly archive the successful exhibit’s results. After sketching ideas and collaborating with my team, I took the initiative to bring the project to life—overseeing everything from initial design concepts to the final lines of code.