Xiaoli Ma

About

Xiaoli Ma is a leading researcher at the intersection of artificial intelligence, cybersecurity, and robotics, with a focus on the robustness of deep reinforcement learning (DRL) systems. Her most-cited work, "Enhanced Adversarial Strategically-Timed Attacks against Deep Reinforcement Learning" (2020), has garnered 25 citations and addresses a critical vulnerability in modern autonomous systems. Ma’s major contribution lies in exposing how strategically timed adversarial perturbations can deceive DRL agents—used in applications from autonomous navigation to robotic arm control—even when those agents are designed for self-adaptation. By demonstrating that attacks need not be constant but can be optimally scheduled to maximize disruption, she has advanced the field’s understanding of security in learning-based control. Her research is pivotal for developing more resilient AI systems, particularly in safety-critical domains. Ma’s work not only highlights the fragility of state-of-the-art neural techniques but also provides a foundation for future defenses, making her a key voice in the ongoing effort to bridge the gap between high-performance learning and real-world reliability.

Research Focus

Key Achievements

Top Papers

1. 1
   Enhanced Adversarial Strategically-Timed Attacks against Deep Reinforcement Learning

   25 citations · 2020

Key Collaborators

Contact & Links