Papers
1
Total Citations
2
H-Index
1
About
Ingook Jang is a researcher advancing the field of robot manipulation through deep reinforcement learning, with a particular focus on modular reward shaping and skill acquisition in continuous action spaces. His most-cited work, "Learning Robot Manipulation based on Modular Reward Shaping" (2020), introduces a framework that decomposes complex robotic tasks into manageable sub-goals, enabling more efficient policy learning. By leveraging modular reward functions, Jang addresses key challenges in sparse reward environments, allowing robots to acquire dexterous manipulation skills with improved sample efficiency. This contribution builds on the broader success of deep reinforcement learning in discrete domains, such as Atari gameplay, and extends its applicability to real-world robotic systems. Though his citation count is still growing, Jang’s work is notable for its practical approach to bridging simulation and reality, offering a pathway for robots to learn intricate tasks like grasping and assembly. His research holds promise for advancing autonomous robotics, particularly in manufacturing and service applications where adaptive, learned behaviors are critical.
Research Focus
Key Achievements
Top Papers
- 1Learning Robot Manipulation based on Modular Reward Shaping2 citations · 2020