Interactive video design for satirical dance Show “SYNOPTICO”. Most of the motion graphics in this piece are built entirely in HTML — using fictional user data to recreate the interfaces and interactions of dating apps and social media, staging a satirical critique of the attention economy.
Particle Theory is an experimental theatre piece grounded in Richard Foreman’s unpublished script of the same title (1973). It stages the performance environment as a 3D software viewport — a cognitive interface — asking what it means to perform and to watch inside a world governed by the logic of a rendering machine.
This three-channel video installation begins in Koreatown, LA — where the artist born and raised in Korea, encountered the layered languages of a diaspora neighborhood: aging signage mixing first-generation immigrants' Korean with broken English, and a newer generation's broken Korean with new English.
Video Projection for Volver, Walt Disney Modular Theater. 4-channel projection design featuring a large-scale laundry screen. Process included screen design, fabrication, 3D mockup, animation, and applied ML model integration.
iF Design Award 2026 Winner for Interior Architecture and Sports/Special Use Interiors.
Commissioned by RIGEND Film to develop AI VFX for aespa's debut album, reaching 166M views. Pipeline developed in ComfyUI.
A solo performance with projectors, about rats and being a rat. It’s a love story.
A mixed-reality film that accuses the outsourcing of our vision to rendering machines, while demonstrating how film language can become the language of technology.
A pipeline that bridges a vision-language model (Gemini) with Segment Anything Model 3 to automatically detect, label, localize, and segment regions in video, exporting per-region masks for downstream use. I named it SAMhera for now.
As Industry–Academia Research Lead, LG Electronics, Seoul — proposed a UX prototype for a file-sorting ML function, adopted as a default feature in the gram Link app for the 24" LG gram Pro series.
A performance exploring the embodied rituals of the Korean bathhouse, drawn from memories I have observed since the age of seven. Though not a single drop of water was used, the space was thick with humidity, created through light and aluminum foil, evoking the dampness and sticky residue of time. By sharing this sensory nostalgia of Korean rituals with dancers from entirely different cultural backgrounds, the work summoned deeply intimate recollections in unexpected ways.
Daily Prototype of poems lively generated, grammatically parsed on the body. Built with TouchDesigner with a local language model.
Designed and delivered an online curriculum, 3D Motion Graphics with Stable Diffusion, introducing innovative workflows that integrate traditional 3D motion graphics with emerging generative models. Developed and demonstrated advanced pipelines using ComfyUI to expand creative control and experimentation. Provided in-depth instruction on ControlNet, IPAdapter, AnimateDiff, and model checkpointing, along with guiding designers and engineers in building custom workflows during the early adoption phase of ComfyUI. Received strong reviews across a range of learners — from technically adaptable designers to engineering students with a design interest.
Richard Foreman (1937–2025) founded the Ontological-Hysteric Theater in New York City in 1968 and spent over five decades building a body of work that refused nearly every convention audiences bring into a theater. His plays do not tell stories; they expose the act of watching.
The writing explores how filmmakers construct and play with spatiality through film language, and how that spatiality (de)constructs narrative. Drawing on the concepts of rochi, raum, and Mnemosyne, alongside examples from experimental theatre and the emerging language of mixed reality — and in dialogue with materials from the 2019 Lisbon Architecture Triennale.
This writing imagines what Harun Farocki and fellow image theorists would say about this fucking situation in terms of image-technology — or rather, what I want to ask them.. while bringing the thoughts from my experience of working on Meta’s SAM 3(Segment Anything Model), and controlNets.
Screen replacing mirrors, semiconductors converting photons into electrons, cameras becoming sensors…
In-Vehicle Entertainment studio leveraging automotive data for personalized content experiences