Hyunwoo Ryu
Papers
5
Total Citations
40
H-Index
3
About
Hyunwoo Ryu is a rising star in robot learning, whose work is redefining how robots understand and interact with the physical world. His research centers on **equivariant deep learning**, specifically leveraging **SE(3) symmetry** to make robotic manipulation dramatically more sample-efficient and robust. Ryu’s major contributions lie in developing novel frameworks that bake geometric priors directly into neural networks. His pioneering work, *Equivariant Descriptor Fields*, introduced SE(3)-equivariant energy-based models for end-to-end visual manipulation, showing how spatial roto-translation equivariance can drastically reduce the number of demonstrations needed for learning. Building on this, his *Diffusion-EDFs* framework (2024, 18 citations) represents a leap forward, combining bi-equivariant denoising generative modeling with SE(3) geometry to handle stochastic human demonstrations with unprecedented fidelity. These contributions are foundational, with his papers accumulating over 40 citations in just a few years. Ryu has also authored a comprehensive tutorial survey on SE(3)-equivariant robot learning, establishing him as a leading voice in the field. His work is not just theoretical; it directly addresses the critical bottleneck of data efficiency in robotics, paving the way for more adaptable and intelligent robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3SE(3)-equivariant Robot Learning and Control: A Tutorial Survey6 citations · 2025
- 4
- 5