Yushi Cheng

Tsinghua University, Zhejiang University

Papers

Total Citations

H-Index

About

Yushi Cheng is a leading researcher in the security of autonomous systems, with a primary focus on adversarial attacks against LiDAR-based perception. Her work exposes critical vulnerabilities in the sensor stacks that underpin self-driving vehicles and robotics. Cheng’s major contribution is the systematic characterization and quantification of physical laser attacks on LiDAR. Her highly cited paper, “PLA-LiDAR: Physical Laser Attacks against LiDAR-based 3D Object Detection in Autonomous Vehicle” (2023, 51 citations), was among the first to demonstrate that an attacker could inject spoofed point clouds using a low-cost laser, causing a vehicle to misclassify or fail to detect real obstacles—a direct threat to safe driving. She extended this work in “Laser-Based LiDAR Spoofing: Effects Validation, Capability Quantification, and Countermeasures” (2024, 5 citations), providing a rigorous framework for evaluating attack effectiveness and proposing defensive strategies. By bridging the gap between theoretical security research and real-world sensor physics, Cheng has established a new research frontier in cyber-physical system security. Her findings are essential reading for anyone working on the safety and robustness of autonomous navigation, and her work directly informs the design of more resilient perception systems.