Papers
2
Total Citations
21
H-Index
2
About
Yu Shen is a leading researcher at the intersection of autonomous systems and multi-robot perception, whose work addresses critical challenges in how machines understand and navigate the physical world. Her primary research areas include adversarial robustness for autonomous driving, cross-view place recognition, and collaborative perception across heterogeneous robot teams. Shen’s most impactful contribution is the development of **Adversarial Differentiable Data Augmentation**, a novel framework that generates realistic, worst-case image degradations to train neural networks for autonomous systems. This work, which has garnered **16 citations**, directly tackles the vulnerability of neural networks to input distortions—a critical safety issue for self-driving cars and drones. In her more recent work on **Visual, Spatial, Geometric-Preserved Place Recognition** (5 citations), Shen pioneered a semantic-based approach that enables aerial and ground robots to reliably identify the same location despite vastly different viewpoints and sensor modalities. This breakthrough is foundational for collaborative search-and-rescue missions. Recognized for bridging the gap between theoretical robustness and practical deployment, Shen’s research is shaping the next generation of resilient, cooperative autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1Adversarial Differentiable Data Augmentation for Autonomous Systems16 citations · 2021
- 2