Yunzhe Tian

About

Yunzhe Tian’s research lies at the intersection of reinforcement learning (RL) and AI security, with a focus on building robust, trustworthy autonomous systems. His most cited work, “Curricular Robust Reinforcement Learning via GAN-Based Perturbation Through Continuously Scheduled Task Sequence” (2022, 12 citations), introduces a novel framework that uses generative adversarial networks to create adaptive perturbations, training RL agents through a continuously scheduled sequence of increasingly challenging tasks. This curricular approach significantly improves the robustness of RL models, particularly for applications in autonomous distributed systems like cooperative robotics. Tian also addresses the critical issue of adversarial vulnerabilities in path planning, as seen in his 2021 paper “A Training-Based Identification Approach to VIN Adversarial Examples in Path Planning” (2 citations), which proposes a method to detect and mitigate adversarial attacks on value iteration networks. By tackling both the performance and security dimensions of AI, Tian’s work contributes to the safe deployment of machine learning in real-world autonomous systems. His research is particularly relevant for students and engineers working on robust RL, adversarial machine learning, and the development of resilient AI for robotics and autonomous navigation.

Research Focus

Key Achievements

Top Papers

1. 1
   Curricular Robust Reinforcement Learning via GAN-Based Perturbation Through Continuously Scheduled Task Sequence

   12 citations · 2022
2. 2
   A Training-Based Identification Approach to VIN Adversarial Examples in Path Planning

   2 citations · 2021

Key Collaborators

Contact & Links