Yiding Jiang

Papers

Total Citations

H-Index

About

Yiding Jiang is a leading researcher in robotics and machine learning, whose work bridges the gap between perception and physical manipulation. His key research areas include robust robot grasping, adversarial learning, and the intersection of computer vision with robotic control. Jiang’s major contribution lies in identifying and formalizing "adversarial grasp objects"—a novel concept inspired by adversarial examples in computer vision. In his highly cited 2019 paper, he demonstrated that learning-based grasp planners, while effective on a wide variety of objects, can be systematically fooled by objects that are physically similar to familiar ones but cause catastrophic failure. This work, with over 30 citations, has profound implications for the safety and reliability of autonomous robotic systems in real-world environments, from manufacturing to service robotics. By exposing these vulnerabilities, Jiang has pushed the field toward more robust and generalizable manipulation algorithms. His research is essential reading for anyone interested in the challenges of deploying deep learning in physical systems, and his insights continue to shape how robots interact with the unpredictable, unstructured world.