Papers
9
Total Citations
158
H-Index
5
About
Deok-Hwa Kim is a robotics researcher whose work bridges the critical gap between robot perception, memory, and autonomous task execution. His primary research areas include visual odometry, biologically inspired episodic memory for robots, and sampling-based motion planning. Kim’s most impactful contribution is his pioneering work on endowing robots with task intelligence through neural models inspired by human cognition. His highly cited 2016 paper on the “Task Intelligence of Robots” (33 citations) introduced a neural model-based mechanism of thought that integrates episodic memory for storing temporal event sequences with online motion planning, enabling robots to reason and act autonomously. This was further advanced in his 2017 work on the “Deep ART Neural Model” (33 citations), which applied biologically inspired episodic memory to help robots perform complex, human-scale tasks. In perception, his 2015 paper on “Visual Odometry Using an RGB-D Sensor and IMU in a Highly Dynamic Environment” (64 citations) remains his most cited work, addressing robust localization in challenging conditions. Kim’s research has earned over 150 total citations, with his work on task intelligence and episodic memory representing a significant step toward truly autonomous service robots capable of learning from experience.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9Large-scale Localization Datasets in Crowded Indoor Spaces2 citations · 2021