
Installation view of Champagne Supernova at BB&M Gallery, 2025.
Champagne Supernova is a single-channel 4K video, 88 minutes in length, composed of eighty-eight one-minute sequences. The artist extracts the iris’s radial patterns using a deep-learning-based AI super-resolution framework (CoZ), then employs them as training data to generate variations of light and colour. The resulting images take in, layer upon layer, textures captured from nature—clouds, waves, rocks, and micro-organisms—forming a vast panorama as if a universe were being born inside the eye. The structure reconfigures once per minute: minute specks of colour ripple and then, at a certain point, congeal into contours like planetary orbits or coastlines, only to divide again like cell membranes and disappear. These recurrences and variations metaphorise the numerical orders humanity has invented to comprehend the cosmos—the eighty-eight constellations officially designated by the International Astronomical Union (IAU) in 1930, the piano’s eighty-eight keys, and the two circles (8–8) that denote infinity. Binding the microscopic structure of the visual organ to a cosmic scale in a single continuous breath, the work demonstrates experientially that the act of looking itself generates an immense terrain. The iris images are rigorously anonymised so that, ultimately, no personal information remains; the private origin of the data is effaced, leaving only traces of pure light. The viewer navigates these shards of light, invited into the universe dormant within the eye, or into the order of rotation and overlap allegorised by the ‘wheels within wheels’ of Ezekiel—the Ophanim. In that encounter, we find ourselves placed in the very position of being seen.


Champagne Supernova No.1 - Libra, 2025. Single-channel 4K video, colour. 1 mins (left)
Champagne Supernova No.4 - Leo, 2025. Single-channel 4K video, colour. 1 mins (right)



Reference book for Champagne Supernova
Artist Jinjoon Lee
Art Director Sun Kim
Technical Director Jinju Kim
AI-assisted Visual Development Doyo Choi
Image Composition & Refinement Shinhyang Park
AI-assisted Video Synthesis Junyeop Park
Researcher Haesoo Han
Technical Director Jinju Kim
AI-assisted Visual Development Doyo Choi
Image Composition & Refinement Shinhyang Park
AI-assisted Video Synthesis Junyeop Park
Researcher Haesoo Han