Yaoyu Hu
Papers
3
Total Citations
41
H-Index
3
About
Yaoyu Hu is a robotics researcher specializing in robot learning, autonomous aerial systems, and physics-based optimization for robotic perception. His work sits at the critical intersection of deep learning and classical robotics, addressing one of the field's most persistent challenges: building systems that generalize reliably across dynamic, real-world environments. Hu's most significant contribution is his involvement in developing **PyPose**, a library designed to bridge the gap between data-driven deep learning and physics-based optimization for robot learning. This work, which has accumulated 35 citations since its 2023 publication, provides the robotics community with a powerful open-source tool that combines the representational strength of neural networks with the generalization capabilities of physics-informed methods — a combination increasingly recognized as essential for robust autonomous systems. Expanding his focus to aerial robotics, Hu co-created the **FIReStereo dataset**, a pioneering forest infrared stereo benchmark enabling depth perception for unmanned aerial systems operating in visually degraded environments such as smoke or darkness. This dataset addresses a critical gap in thermal imaging research for autonomous flight. Through these contributions, Hu has established himself as an emerging voice in robust, safety-critical robot perception research.
Research Focus
Key Achievements
Top Papers
- 1PyPose: A Library for Robot Learning with Physics-based Optimization35 citations · 2023
- 2
- 3PyPose: A Library for Robot Learning with Physics-based Optimization3 citations · 2022