Lantao Yu

Papers

Total Citations

H-Index

About

Lantao Yu is a researcher at the intersection of reinforcement learning, robotics, and medical imaging, with a focus on enabling intelligent systems to learn from complex, real-world environments. His most influential work, "Adversarial Inverse Reinforcement Learning With Self-Attention Dynamics Model" (2021, 32 citations), addresses the fundamental challenge of learning optimal policies from expert demonstrations when reward functions are difficult to specify. By integrating self-attention mechanisms into adversarial inverse reinforcement learning, Yu's approach improves robustness against stochastic environments—a critical advancement for applications ranging from autonomous driving to surgical robotics. In the medical domain, Yu's "Intuition-guided Reinforcement Learning for Soft Tissue Manipulation with Unknown Constraints" (2024) pioneers robotic decision-making under dynamic, uncertain conditions, tackling the previously unsolved problem of grasping and deforming soft tissues without prior knowledge of constraints. Most recently, his work "Ultra-Malin: Robotic Ultrasound Mapping and Localization via Implicit Neural Representation" (2025) introduces a novel framework that overcomes the limitations of traditional ultrasound image localization, enabling robots to map and navigate anatomical structures without relying on fixed positional features. Together, these contributions demonstrate Yu's commitment to bridging theoretical reinforcement learning advances with practical robotic systems that can operate safely and effectively in unstructured, real-world environments.