Seungyong Lee

My research interests lie in computer vision, particularly in both understanding and generating complex visual worlds. I am interested in visual recognition as a way to perceive the world, and in generative models as a means to recreate and manipulate it. I am also passionate about large-scale model training and scalable inference systems from an engineering perspective. At NXN Labs, I have led the development of virtual try-on models and image generation frameworks, taking ownership of the full pipeline—from problem definition and data collection to model design, training, and deployment. I studied Mathematics and Electrical Engineering at KAIST, where I am currently on a leave of absence.

Email  /  CV  /  Github  /  LinkedIn

We introduced a unified and scalable diffusion transformer for jointly learning virtual try-on and try-off, enabling robust garment–target correspondence.

We proposed a unified diffusion transformer that performs both virtual try-on and try-off, achieving state-of-the-art results on both tasks.

We achieve high fidelity text-driven fashion style editing in a compute-efficient manner by leveraging a generative human prior.

NXN Labs
AI Research Engineer, Founding Member (Jul. 2023 ~)

Led the end-to-end development of foundation models for fashion image generation and virtual try-on. Actively contributed to the entire pipeline—from problem definition to data acquisition and processing, model design, training, evaluation, and deployment—demonstrating ownership across research and engineering.

Lunit
AI Research Scientist Intern (Jan. 2023 ~ Jul. 2023)

Worked on medical image recognition using computer vision techniques. Focused on tumor cell detection and segmentation in pathological images, leveraging Vision Transformers to improve accuracy and robustness in challenging clinical scenarios.

Koh-yong Technology R&D Center
AI Engineer Intern (Aug. 2020 ~ Feb. 2021)

Participated in an anomaly detection project for inspecting defects on semiconductor substrates. Gained hands-on experience in building data labeling tools, training deep learning models, and deploying real-time inference systems.

The template is from Jon Barron. Thank you for sharing!