Cunjun Yu
Hi! I'm a PhD student at the National University of Singapore, under the guidance of Professor David Hsu. I aim to build robots capable of teaching humans a variety of skills, such as collaborating effectively, writing Chinese calligraphy, playing air hockey, and driving. I also work on planning under dynamic environments, for example, autonomous driving and human-robot collaboration.
I am deeply grateful for the opportunity to have previously interned at 
NVIDIA Research and 
Toyota Research Institute, where I gained invaluable experiences.
Github Google Scholar LinkedIn
cunjun dot yu at gmail dot com
News (Expand)
Publications ( show selected / show all by date )
“Stack It Up!": 3D Stable Structure Generation from 2D Hand-drawn Sketch
Conference on Robot Learning (CoRL), Oral Presentation, 2025
GSON: A Group-based Social Navigation Framework with Large Multimodal Model
Robotics and Automation Letters (RA-L), 2025
Robi Butler: Remote Multimodal Interactions with Household Robot Assistant
International Conference on Robotics and Automation (ICRA), 2025
InsActor: Instruction-driven Physics-based Characters
Neural Information Processing Systems (NeurIPS), 2023
DiffMimic: Efficient Motion Mimicking with Differentiable Physics
International Conference on Learning Representations (ICLR), 2023
DaXBench: Benchmarking Deformable Object Manipulation with Differentiable Physics
International Conference on Learning Representations (ICLR), Notable Top 5%, 2023
What Truly Matters in Trajectory Prediction for Autonomous Driving?
Neural Information Processing Systems (NeurIPS), 2023
PDF • Project Page • Code
Balanced MSE for Imbalanced Visual Regression
Conference on Computer Vision and Pattern Recognition (CVPR), Oral Presentation, 2022
PDF • Code • Project Page
INVIGORATE: Interactive Visual Grounding and Grasping in Clutter
Robotics: Science and Systems (RSS), 2021
Spatio-Temporal Graph Transformer Networks for Pedestrian Trajectory Prediction
European Conference on Computer Vision (ECCV), 2020
PDF • Code • Project Page • Video
Balanced Meta-Softmax for Long-Tailed Visual Recognition
Neural Information Processing Systems (NeurIPS), 2020
Arxiv • Code • Project Page
MessyTable: Instance Association in Multiple Camera Views
European Conference on Computer Vision (ECCV), 2020
Arxiv • Code • Project Page
Siamese Convolutional Neural Network for Sub-millimeter-accurate Camera Pose Estimation and Visual Servoing
International Conference on Intelligent Robots and Systems (IROS), 2019