Seunghoi Kim

University College London

Papers

2

Total Citations

6

H-Index

2

About

Seunghoi Kim is a rising researcher at the intersection of computer vision, robotics, and medical imaging, with a focus on advancing 3D perception for surgical and autonomous systems. His work tackles fundamental challenges in depth estimation and point cloud segmentation, aiming to bridge the gap between general-purpose foundation models and specialized, high-stakes applications like robotic-assisted surgery. In his 2025 paper "DARES," Kim pioneered a self-supervised Vector-LoRA approach to adapt the powerful Depth Anything Model (DAM) for endoscopic surgery, addressing the critical need for accurate 3D reconstruction in the operating room. This work, already garnering 4 citations shortly after release, demonstrates his ability to leverage large-scale models while overcoming the constraints of limited surgical data. Earlier, in "AGCN" (2021), he introduced an Adversarial Graph Convolutional Network to tackle the persistent challenge of object boundary ambiguity in 3D point cloud segmentation—a problem central to medical robotics and autonomous driving. Though early in his career, Kim’s contributions signal a promising trajectory in making deep learning models more robust, adaptable, and clinically relevant.

Research Focus

Key Achievements

2
H-Index
2
Papers
6
Total Citations
3
Avg Citations/Paper
🏆 Most Cited Paper
DARES: Depth Anything in Robotic Endoscopic Surgery with Self-supervised Vector-LoRA of the Foundation Model
4 citations · 2025
📈 Most Prolific Year: 2025 (1 Papers)
🤝 Key Collaborators: 14
🏛 Institutions: University College London

Top Papers

  1. 1
  2. 2

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago