Papers
13
Total Citations
329
H-Index
9
About
Inwook Shim is a leading roboticist whose research bridges perception, control, and autonomy for robots operating in challenging, unstructured environments. His most impactful work stems from his integral role on Team KAIST during the DARPA Robotics Challenge (DRC) Finals, where his contributions to the robot system and control strategy of DRC-HUBO+ (127 citations) helped the team secure victory in the disaster-response competition. This success was underpinned by his development of a robust software system for operating humanoid robots under degraded conditions (15 citations). Beyond humanoid control, Shim has made foundational contributions to vision for outdoor robotics. He pioneered a gradient-based camera exposure control method (55 and 46 citations) that ensures reliable image feature extraction even under harsh lighting, a critical capability for autonomous navigation. His work extends to 3D perception, including urban structure classification using NDT grids (25 citations) and a geometry-guided deep network for point cloud registration (15 citations). More recently, he has advanced self-supervised learning for off-road traversability estimation (10 citations), enabling robots to navigate complex terrain without manual labels. With over 300 total citations, Shim’s work is essential reading for researchers in field robotics, autonomous navigation, and robot perception.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Gradient-Based Camera Exposure Control for Outdoor Mobile Platforms46 citations · 2018
- 4
- 5
- 6Geometry Guided Network for Point Cloud Registration15 citations · 2021
- 7Self-Supervised 3D Traversability Estimation With Proxy Bank Guidance10 citations · 2023
- 8Geometric-featured voxel maps for 3D mapping in urban environments10 citations · 2011
- 9
- 10Autonomous homing based on laser-camera fusion system7 citations · 2012