Houqing Wang

Papers

Total Citations

H-Index

About

Houqing Wang is a leading researcher in 3D perception and adversarial machine learning, with a focus on securing deep learning models for critical real-world applications like autonomous driving and robotics. Their most-cited work, "The Art of Defense: Letting Networks Fool the Attacker" (2023, 19 citations), addresses a fundamental vulnerability in 3D object classifiers that rely on point cloud data. Wang demonstrates how these state-of-the-art (SOTA) systems can be deceived by adversarial attacks, and proposes innovative defense strategies that turn the tables on attackers—essentially teaching networks to mislead adversaries rather than simply resist them. This contribution is pivotal for ensuring the reliability of perception systems in safety-critical environments. By bridging the gap between 3D vision and adversarial robustness, Wang’s research has garnered attention from both academia and industry, laying groundwork for more resilient autonomous systems. Their work stands out for its practical impact, offering a proactive approach to security that could redefine how we design trustworthy AI for the physical world.