Xinrun Li

Robert Bosch (United States), Robert Bosch (China)

Papers

Total Citations

H-Index

About

Xinrun Li is a researcher advancing the frontiers of autonomous driving perception and sensor performance evaluation. Their work centers on two critical areas: lane topology extraction for mapless driving and systematic sensor visibility assessment. Li’s most impactful contribution, “Chameleon: Fast-Slow Neuro-Symbolic Lane Topology Extraction” (2025, 5 citations), introduces a novel neuro-symbolic framework that combines fast neural perception with slow symbolic reasoning to infer complex lane relationships—such as determining permissible turning maneuvers—a key challenge for safe, mapless autonomous navigation. This work bridges the gap between data-driven perception and rule-based reasoning, offering a scalable solution for real-world driving scenarios. In parallel, Li’s earlier research, “Sensor Visibility Estimation: Metrics and Methods for Systematic Performance Evaluation and Improvement” (2022, 3 citations), establishes foundational metrics for quantifying where sensors can or cannot measure, directly enhancing functional safety in automotive, robotics, and smart infrastructure systems. By providing rigorous evaluation methods, this work enables engineers to identify and mitigate sensor blind spots, improving system reliability. Li’s research is notable for its practical impact on safety-critical applications, demonstrating a clear trajectory from sensor-level analysis to high-level scene understanding.