About

Gyuho Eoh is a leading researcher in multi-robot systems and autonomous navigation, with a focus on cooperative object transportation and collision avoidance. His work bridges classical control theory and modern deep reinforcement learning (DRL), as demonstrated by his highly cited paper "Multi-robot cooperative formation for overweight object transportation" (29 citations), which pioneered formation-based approaches for handling heavy loads. Eoh's recent contributions include "Cooperative Object Transportation Using Curriculum-Based Deep Reinforcement Learning" (21 citations), where he introduced automatic curriculum design to solve the sparse reward problem in DRL-based transportation tasks. His research also addresses critical safety challenges, such as in "Analytic collision anticipation technology considering agents' future behavior" (15 citations), which predicts when and where collisions will occur by modeling agents' future trajectories. Eoh's work on faulty robot rescue (12 citations) and reactive free space estimation (9 citations) further showcases his versatility in developing robust, decentralized solutions for dynamic environments. His innovative use of variational Bayesian methods for condition-invariant feature extraction (7 citations) highlights his commitment to long-term autonomy. With over 100 total citations, Eoh's research continues to shape the future of cooperative robotics.

Research Focus

Key Achievements

7
H-Index
14
Papers
126
Total Citations
9
Avg Citations/Paper
🏆 Most Cited Paper
Multi-robot cooperative formation for overweight object transportation
29 citations · 2011
📈 Most Prolific Year: 2021 (4 Papers)
🤝 Key Collaborators: 22
🏛 Institutions: Seoul National University, Chungbuk National University, Seoul National University of Science and Technology, Korea University of Technology and Education

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago