Minjian Xin

About

Minjian Xin is a robotics researcher whose work centers on deep reinforcement learning (DRL) for complex manipulation tasks, particularly those involving long horizons and sparse rewards. His major contribution is a novel framework that leverages base controllers—predefined, low-level motion primitives—to guide exploration in DRL, effectively overcoming the inefficiency that plagues traditional sparse-reward settings. By integrating these base controllers, his approach enables robots to learn intricate, multi-step behaviors, such as assembly or tool use, without requiring dense reward signals. His most-cited paper (2022, 23 citations) demonstrates this method on robotic manipulator tasks, achieving significant improvements in sample efficiency and task success rates. This work has been recognized for bridging the gap between classical control and modern learning-based methods, offering a practical pathway for deploying DRL in real-world industrial automation. Xin’s research is particularly impactful for students and engineers seeking to apply reinforcement learning to physical systems, as it addresses a core bottleneck in robotics: learning from sparse feedback. His ongoing work continues to push the boundaries of autonomous manipulation, making him a rising voice in the field.

Research Focus

Key Achievements

Top Papers

1. 1
   Learning of Long-Horizon Sparse-Reward Robotic Manipulator Tasks With Base Controllers

   23 citations · 2022

Key Collaborators

Contact & Links