Seekr is a desktop research prototype combining Gaussian Splatting, an AI-generated world model, and natural-language interaction to explore human and AI spatial representations. I designed and implemented an AI perception pipeline that transforms human visual observations into navigable representations through camera-based image capture, object detection, and real-time trajectory visualization. This project became the starting point of my current research on personalized world models and spatial cognition.

Can the felt sense of presence survive the body's absence in virtual space? In this research-through-design VR project (Unity, Quest 3), users enter a scanned digital twin of their own surroundings and explore it through guided, touch-oriented interaction; across an in-progress exhibition and user testing, participants found that bodily absence heightened their awareness of presence rather than diminishing it. The work reframes touch as a means of constructing presence rather than simulating realism, and opens testable questions about how tactile and bodily cues shape presence in VR.

At what point do people recognize themselves inside an AI-generated image? In this interactive installation, participants' body movement drove real-time generative visuals, letting us probe the threshold of self-recognition in digital imagery. We found that both movement and image content shifted recognition sensitivity, and are extending the work to new physical environments to test how context shapes self-perception. Published at ACM IMX 2025.

Does the body register what we read below awareness? During an Immersion Lab residency, I measured content-dependent micro-variations in hand pressure while participants read on-screen—behavioral evidence of unconscious bodily response to media content.

Metal Applicator is a practice-based research project exploring how applying digital materials can become an embodied interaction rather than an instantaneous command. Combining Unity, Arduino, and physical prototyping, I designed and built a system that transforms digital material assignment into a tangible, bodily action. The project demonstrates how physical effort can reshape our experience of digital materiality and interaction in 3D environments.

Spolia Lounge is a physical installation exploring the continuous translation between digital and physical matter. Photographs of the exhibition space were transformed into textile patterns and re-materialized onto garments, furniture, books, and everyday objects within the same environment. By enabling visitors to recognize familiar objects through their abstracted material traces, the installation makes the ongoing exchange between digital and physical space tangible.

Does learner agency deepen engagement in immersive learning, or is guided repetition enough? Building on Kukshinov's work on VR and perspective-taking, I designed a VR Korean-learning experience (Unity, Quest 3) structured as three interactive activities with an AI NPC guide ("Yunsun") that answers learners' own interest-driven questions from a curated knowledge bank rather than scripted prompts. The project reframes immersive learning around agency and inquiry, opening testable questions about how agency and conversational grounding affect engagement and perspective-taking in VR.