I am Weihao Yuan (原玮浩), currently an algorithm expert at Alibaba. I obtained my PhD in Robotics Institute, Department of Electronic & Computer Engineering, Hong Kong University of Science and Technology (HKUST), supervised by Prof. Michael Yu Wang and co-supervised by Prof. Qifeng Chen. I visited the Robotics, Perception and Learning Lab, Royal Institute of Technology (KTH) for 6 months, supervised by Prof. Danica Kragic . I received my B.S. degree from College of Control Science and Engineering, Zhejiang University, majoring in Automation and minoring in Finance.
My research areas include Deep Reinforcement Learning, Robotic Motion Planning, 3D Reconstruction, 3D Generation, Human Motion, Video Generation
News:
2024-07-30
One paper accepted to SIGGRAPH Asia & ACM Transactions on Graphics (TOG):
- MVImgNet2.0: A Larger-scale Dataset of Multi-view Images
2024-07-10
One paper accepted to IEEE Robotics and Automation Letters (RA-L):
- OV9D: Open-Vocabulary Category-Level 9D Object Pose and Size Estimation
2024-07-05
Three papers accepted to European Conference on Computer Vision (ECCV 2024):
- Freditor: High-Fidelity and Transferable NeRF Editing by Frequency Decomposition
- An Optimization Framework to Enforce Multi-View Consistency for Texturing 3D Meshes Using Pre-Trained Text-2-Image Models
- High-Fidelity 3D Textured Shapes Generation by Sparse Encoding and Adversarial Decoding
2024-02-27
Three papers accepted to IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2024):
- IPoD: Implicit Field Learning with Point Diffusion for Generalizable 3D Object Reconstruction from Single RGB-D Images Highlight
- RichDreamer: A Generalizable Normal-Depth Diffusion Model for Detail Richness in Text-to-3D Highlight
- GPLD3D: Latent Diffusion of 3D Shape Generative Models by Enforcing Geometric and Physical Priors Oral
2023-03-06
One paper accepted to IEEE Robotics and Automation Letters (RA-L):
- DRO: Deep Recurrent Optimizer for Video to Depth
2023-01-21
Two papers accepted to International Conference on Learning Representations (ICLR 2023):
- 3D Former: Monocular Scene Reconstruction with 3D SDF Transformers
- DIM-SLAM: DENSE RGB SLAM WITH NEURAL IMPLICIT MAPS
2022-03-03
Two papers accepted to IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2022):
- NeW CRFs: Neural Window Fully-connected CRFs for Monocular Depth Estimation
- RCP: Recurrent Closest Point for Scene Flow Estimation on 3D Point Clouds Oral
2021-07-01
One paper accepted to IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021):
- Stereo Matching by Self-supervision of Multiscopic Vision
2020-07-01
Two papers accepted to IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2020):
- Self-supervised Object Tracking with Cycle-consistent Siamese Networks
- Multi-Object Rearrangement with Monte Carlo Tree Search: A Case Study on Planar Nonprehensile Sorting
2020-01-24
One paper accepted to IEEE Robotics and Automation Letters (RA-L):
- MFuseNet: Robust Depth Estimation with Learned Multiscopic Fusion
2020-01-22
One paper accepted to IEEE International Conference on Robotics and Automation (ICRA 2020):
- MFuseNet: Robust Depth Estimation with Learned Multiscopic Fusion
2019-09-09
One paper accepted to Conference on Robot Learning (CoRL 2019) as oral presentation! (Acceptation rate: 5.3%):
- Toward Learning to Detect and Predict Contact Events on Vision-based Tactile Sensor
2019-06-06
One paper accepted to Robotics and Autonomous Systems (RAS):
- End-to-End Nonprehensile Rearrangement with Deep Reinforcement Learning and Simulation-to-Reality Transfer
2019-02-01
One paper accepted by IEEE International Conference on Robotics and Automation (ICRA 2019), May 20-24, 2019, Montreal, Canada:
- Reinforcement Learning in Topology-based Representation for Human Body Movement with Whole Arm Manipulation
2018-03-27
I have begun the visiting in KTH Royal Institute of Technology, Storkholm, Sweden, as a visiting researcher for next 6 months, working with Prof. Danica Kragic and Dr. Johannes Andreas Stork.
2018-01-12
One paper accepted to IEEE International Conference on Robotics and Automation (ICRA 2018), May 21-25, 2018, Brisbane, Australia.
- Rearrangement with Nonprehensile Manipulation Using Deep Reinforcement Learning