about | Seungwoo Kum

Korea Electronics Technology Institute

Seongnam, Gyeonggi-do, Korea

I am the Director of the AI Data & Security Research Center at the Korea Electronics Technology Institute (KETI), where I lead research on AI data infrastructure, security, and edge AI systems.

My work bridges fundamental AI research and practical deployment — from designing data pipelines for agricultural and industrial applications to building secure edge AI architectures for autonomous vehicles and IoT environments. I also coordinate international collaborative research with partners in Europe, Spain, Czech Republic, and Canada.

Research Interests

AI Data Platforms & Data Quality
Security for AI Systems
Edge AI & On-device Intelligence
Industrial AI Agents
International AI Collaboration (EU, Korea-Spain, Korea-Canada)

Current Focus

I am currently leading new initiatives in industrial AI agent technology (2026–) and Korea-Spain AI collaboration, while continuing to develop AI talent through the national AI workforce development program.

Feel free to reach out if you are interested in collaboration or joint research.

selected projects

Generative AI Talent Development Program

Workforce program for generative AI; KETI's research develops Vision-Language-Action (VLA) models for robot manipulation

KRESIP — Korea–Spain Smart-Farm AI Platform

AI-based precision-agriculture platform jointly developed by Korea and Spain

selected publications

Hierarchical Attention-based Graph Neural Network with Relevance-driven Pruning

Seungwoo Kum

arXiv preprint arXiv:2605.09308, 2026

arXiv Bib

@article{kum2026hierarchical,
  title = {Hierarchical Attention-based Graph Neural Network with Relevance-driven Pruning},
  author = {Kum, Seungwoo},
  journal = {arXiv preprint arXiv:2605.09308},
  year = {2026},
}

Novel method for crop growth tracking with deep learning model on an Edge Rail Camera

Seungwoo Kum, Jaewon Moon, Sangtae Oh, and 7 more authors

Computers and Electronics in Agriculture, 2025

DOI Bib

@article{kum2025railcamera,
  title = {Novel method for crop growth tracking with deep learning model on an Edge Rail Camera},
  doi = {10.2139/ssrn.4583976},
  author = {Kum, Seungwoo and Moon, Jaewon and Oh, Sangtae and Suh, H. K. and Park, H. and Sim, H. S. and Jo, J. S. and Kim, S. K. and Choi, S. and others},
  journal = {Computers and Electronics in Agriculture},
  volume = {230},
  pages = {109816},
  year = {2025},
}

Edge AI framework for large scale smart agriculture

Seungwoo Kum, Sangtae Oh, and Jaewon Moon

In 2024 27th Conference on Innovation in Clouds, Internet and Networks (ICIN), 2024

DOI Bib

@inproceedings{kum2024edgeai,
  title = {Edge AI framework for large scale smart agriculture},
  doi = {10.1109/icin60470.2024.10494451},
  author = {Kum, Seungwoo and Oh, Sangtae and Moon, Jaewon},
  booktitle = {2024 27th Conference on Innovation in Clouds, Internet and Networks (ICIN)},
  year = {2024},
}

Optimization of edge resources for deep learning application with batch and model management

Seungwoo Kum, Sangtae Oh, J. Yeom, and 1 more author

Sensors, 2022

DOI Bib

@article{kum2022sensors,
  title = {Optimization of edge resources for deep learning application with batch and model management},
  doi = {10.3390/s22176717},
  author = {Kum, Seungwoo and Oh, Sangtae and Yeom, J. and Moon, Jaewon},
  journal = {Sensors},
  volume = {22},
  number = {17},
  pages = {6717},
  year = {2022},
}

see all publications →

news

May 23, 2026	Posted a new update on LinkedIn.
Aug 19, 2026	At UAI 2026: My Pipeline Was Far Too Sure of Itself new post
Jul 23, 2026	White Paper: Advancing AI Safety and Security new post
Jul 21, 2026	Three Failures Before a Knowledge Base That Held new post
Jul 13, 2026	Running AI Inside a Trusted Execution Environment new post