Yumin Chen

National Tsing Hua University, Xiamen University of Technology

Papers

Total Citations

H-Index

About

Yumin Chen’s research lies at the intersection of computer vision and robotics, with a primary focus on visual place recognition and sim-to-real transfer for robotic control. In their seminal work “Virtual-to-Real: Learning to Control in Visual Semantic Segmentation” (2018, 69 citations), Chen tackled the critical challenge of bridging the reality gap between synthetic training data and real-world visual environments—a bottleneck for deploying learned robotic policies safely. By leveraging semantic segmentation as a shared representation, this work enabled robots to transfer control policies from simulation to physical platforms without costly real-world data collection. Chen also made foundational contributions to place recognition through the comprehensive survey “Place Recognition: An Overview of Vision Perspective” (2018, 26 citations), which systematically reviewed decades of progress in recognizing locations from images—a core capability for autonomous navigation and augmented reality. This overview remains a key reference for researchers entering the field. Chen’s work is notable for addressing practical deployment challenges in robotics, combining rigorous theoretical analysis with real-world validation. Their research continues to influence how robots perceive and navigate complex environments, advancing both the safety and autonomy of intelligent systems.