Okdalto
한국어

About

Hello, I’m Okdalto.

I’m a computer graphics researcher who explores visual experiments through deep learning and computer graphics. What interests me most is the intersection of technology and art. In particular, I’m drawn to the ways computers generate images, and to the meanings that emerge from those processes. What may look like an error or artifact to someone else often feels to me like the beginning of a new visual style.

I studied Visual Design at Kookmin University, completed my master’s degree at the KAIST Visual Media Lab, and began my Ph.D. in the same lab. Midway through, I co-founded Anigma Technologies with my labmates and worked there for about three years; I’ve since returned to continue my Ph.D., while serving as an AI advisor at IICOMBINED, the company behind Gentle Monster.

I also care deeply about teaching. Rather than leaving AI as something mysterious, I think it matters to make it something people can take apart and touch. That’s why I share my process and code in the open, and why I’ve kept running workshops that lower the barrier to learning. I’ve given guest lectures on AI at universities including the Korea National University of Arts, Kookmin University, Seoul National University, Sogang University, Sungshin Women’s University, and Soongsil University, and starting in fall 2026, I teach an AI art practice course at the College of Fine Arts, Seoul National University.

My research has covered facial animation retargeting, audio-driven speech animation, and neural rendering, while my creative work spans AI VFX for music videos and interactive media art installations. To me, research and art are ultimately the same job: taking apart the way computers make images, and finding new formal possibilities inside.

On this blog, I plan to write about the process behind my works, thoughts that come up along the way, and occasionally technical tutorials. My goal is to explain even difficult topics as clearly and accessibly as possible.

Works

  • ENHYPEN ‘THE SIN : VANISH’ Chapter 2 video, computer vision (2025)
  • DOECHII Korean-name reveal party video, Universal Music Korea (2025) — making-of
  • suiimeetsyou ‘G e t s u yo b i’ music visualizer (2025) — making-of
  • IRENE ‘Like A Flower’ music video AI VFX (2024) — making-of
  • Sokodomo ‘E-Girl (Feat. So!YoON!)’ music video graphics (2024)
  • pH-1 ‘U TELL ME (Feat. P-Lo)’ music visualizer (2023)
  • Leem & Lee Do-yeol ‘여우 (Fox)’ music video visualizer (2023)
  • Woo Won-jae ‘Ransome’ music video AI VFX (2023) — making-of
  • JO1 ‘TIGER’ music video AI development (2023)
  • BTS × Louis Vuitton editorial media art, Vogue Korea · GQ Korea January issue (2022)
  • Leesuho ‘Lights (feat. Kim Hanjoo)’ music video, director (2021)
  • p5.js Korean homepage background artwork (2020)

Exhibitions

  • ‘Park Seo-Bo x LG OLED: Colors Drawn from Nature’, Frieze Seoul, LG OLED Lounge (2025) — technical director
  • ‘Neural Network’, Design Korea, COEX, Seoul (2024) — making-of
  • ‘Visualized Convolutional Neural Networks with MNIST’, CVPR AI Art Gallery, Seattle (2024) — making-of
  • ‘100: One Hundred Self-Introductions’, eyesmag, piknic, Seoul (2023) — video production
  • ‘Henri Matisse’ special exhibition, Hangaram Art Museum, Seoul Arts Center (2021) — machine learning development
  • ‘Manifold’, Game & Art: Auguries of Fantasy, Daejeon Museum of Art (2021)
  • Yiyun Kang ‘Proxemics: Drawing Social Bubble’, Paradise Walk (2021) — technical director
  • ‘CURVEilance: Multi-User Based Interactive Media Wall’, SIGGRAPH Asia Art Gallery (2020)
  • ‘The Skin’, ACM SIGGRAPH Art Gallery (2020)
  • ‘RE-VIS-IT’, Je Baak’s solo exhibition, Hyundai Motorstudio Seoul (2019) — technical director

Research