Jinyeob Kim

About

Jinyeob Kim is a rising researcher in robotics, specializing in socially aware navigation and human-robot interaction. His work focuses on enabling robots to move safely and naturally alongside humans, particularly in dynamic, crowded environments. Kim’s major contributions include developing a **Transformable Gaussian Reward Function** for deep reinforcement learning, which significantly improves a robot’s ability to navigate socially—a paper that has already garnered 8 citations since its 2024 publication. He has also advanced **human-to-robot handover** using reinforcement learning, integrating adaptive control to allow anthropomorphic grippers to safely exchange diverse objects with people. Additionally, Kim introduced **belief-aided navigation** using Bayesian reinforcement learning, a novel approach that helps robots anticipate and avoid humans in blind spots—addressing a critical gap in practical, real-world deployment where sensor coverage is limited. His work is notable for bridging the gap between theoretical socially aware navigation and practical, sensor-constrained applications. With multiple high-impact papers published in 2024 alone, Kim is establishing himself as a key contributor to the next generation of human-aware robotic systems.

Research Focus

Key Achievements

Top Papers

1. 1
   Transformable Gaussian Reward Function for Socially Aware Navigation Using Deep Reinforcement Learning

   8 citations · 2024
2. 2
   Human-to-Robot Handover Based on Reinforcement Learning

   2 citations · 2024
3. 3
   Belief-Aided Navigation using Bayesian Reinforcement Learning for Avoiding Humans in Blind Spots

   2 citations · 2024

Key Collaborators

Contact & Links