Jiqiang Liu
Papers
2
Total Citations
14
H-Index
2
About
Jiqiang Liu is a leading researcher at the forefront of artificial intelligence security and robust reinforcement learning. His work critically addresses the vulnerabilities in autonomous systems, particularly how adversarial examples—subtly manipulated inputs—can deceive machine learning models and cause catastrophic failures in real-world applications like path planning for autonomous vehicles. Liu’s major contribution lies in developing novel defenses against these threats. He pioneered "Curricular Robust Reinforcement Learning," a groundbreaking approach that uses GAN-based perturbations to train agents through a continuously scheduled sequence of increasingly difficult tasks. This method significantly enhances the resilience of reinforcement learning models, which are essential for autonomous distributed systems such as cooperative robotics. While his most-cited paper has garnered 12 citations, his broader impact is defined by his proactive approach to AI safety, addressing a critical bottleneck in the deployment of intelligent systems. By fortifying the learning process itself against malicious inputs, Liu is helping to build a more secure and reliable foundation for the next generation of autonomous technologies.
Research Focus
Key Achievements
Top Papers
- 1
- 2