|
Peidong LiuAssistant Professor
Spatial Intelligence and Robotics Lab Welcome to follow our latest research on: Twitter, Github, Zhihu !! |
About
Dr. Peidong Liu is currently an Assistant Professor of Computer Science at Westlake University, Hangzhou, China. He got his PhD degree from the Department of Computer Science at ETH Zurich, under the supervision of Prof. Marc Pollefeys. He got both his master and bachelor degrees in Electrical Engineering from National University of Singapore. His research bridges 3D computer vision and robotics, focusing on visual spatial intelligence (SpatialAI) - developing algorithms that enable machines to perceive, understand, and interact with their 3D environment through visual data. Details are as follows:
- Perception: 3D/4D reconstruction and generation;
- Reasoning: 3D scene understanding and reasoning;
- Action: 3D vision for robot;
Recent news
| Jun 18, 2026 | Our new paper “Walking in the Implicit: Interactive World Exploration via Neural Scene Representation” is accepted by ECCV 2026. Congratulations to Zhiqi, Chengrui, Zhenhua and all other co-authors! |
|---|---|
| Jun 17, 2026 | Our new paper “Event3R: Asynchronous-to-Global 3D Reconstruction from Event Camera via Spatial-Temporal Feature Aggregation” is accepted by IROS 2026. Congratulations to Jian, Haotian and all other co-authors! |
| Jun 3, 2026 | Congratulations to my PhD students, Peng Wang and Yunhao Li, on successfully defending their PhD theses and graduating. Peng is going to join RobbyAnt (an embodied AI company under Ant Group), and Yunhao is going to join The University of Hong Kong as a Postdoc. |
| Jan 31, 2026 | 1 papers accepted by ICRA-2026. Congratulations to Huang Jian, Haotian and Xinhao! |
| Jan 27, 2026 | 2 papers accepted by ICLR-2026. Congratulations to Yiming, Bangyan and all other co-authors! |
Current group members
Senior researchers:
- Dongxu Wei: (Joined in June 2024) PhD from Zhejiang University
PhD students:
- Qi Xu (2026 fall intake): master from Wuhan University
- Haotian Shen (2026 fall intake): bachelor from Westlake University
- Zhangchi Huang (2025 fall intake): bachelor from Xi'an Jiaotong University (Qian Xuesen Honors College)
- Zhenhua Du (2025 fall intake): bachelor and master from National University of Defense Technology
- Chengrui Dong (2024 fall intake): bachelor student from Northwestern Polytechnical University
- Yiming Chen (2024 fall intake): bachelor student from Tongji University
- Lingzhe Zhao (2024 fall intake): bachelor from Tongji University
- Bangyan Liao (2023 fall intake): bachelor from Hunan University
- Jian Huang (2022 fall intake): bachelor and master from Xi'an Jiao Tong University
- Zhiqi Li (2022 fall intake): bachelor from Shanghai Jiao Tong University
Internships and visiting students:
- Xiang Liu: master student from ETH Zurich
- Chi Wang: master from Xidian University
- Yaohui Li: master from Northeastern University
- Wenpu Li: bachelor student from Guangdong University of Technology
- Xinhao Lou: bachelor student from Westlake University
Group Alumni
Selected publications
For more details, please refer to my Google Scholar page.
-
Walking in the Implicit: Interactive World Exploration via Neural Scene RepresentationIn European Conference on Computer Vision (ECCV) 2026
-
Event3R: Asynchronous-to-Global 3D Reconstruction from Event Camera via Spatial-Temporal Feature AggregationIn International Conference on Intelligent Robots and Systems (IROS) 2026
-
Reasoning in Space via Grounding in the WorldIn International Conference on Learning Representations (ICLR) 2026
-
Neural Predictor-Corrector: Solving Homotopy Problems with Reinforcement LearningIn International Conference on Learning Representations (ICLR) 2026
-
SCIGaussian-D: Dynamic Scene Reconstruction from a Single Snapshot Compressive ImageIn International Conference on 3D Vision (3DV) 2026
-
ED-SLAM: Event-Depth Gaussian Splatting SLAMIn International Conference on Robotics and Automation (ICRA) 2026
-
SIU3R: Simultaneous Scene Understanding and 3D Reconstruction Beyond Feature AlignmentIn Neural Information Processing Systems (NeurIPS) 2025Spotlight paper, (acceptance ratio: 3.1%)
-
Styl3R: Instant 3D Stylization Reconstruction from Arbitrary Scenes and StylesIn Neural Information Processing Systems (NeurIPS) 2025
-
E-MoFlow: Learning Egomotion and Optical Flow from Event Data via Implicit RegularizationIn Neural Information Processing Systems (NeurIPS) 2025
-
MBA-SLAM: Motion Blur Aware Dense Visual SLAM with Radiance Fields RepresentationTransactions on Pattern Analysis and Machine Intelligence (T-PAMI) 2025
-
CasualHDRSplat: Robust High Dynamic Range 3D Gaussian Splatting from Casually Captured VideosIn ACM International Conference on Multimedia (MM) 2025
-
Omni-Scene: Omni-Gaussian Representation for Ego-Centric Sparse-View Scene ReconstructionIn Computer Vision and Pattern Recognition (CVPR) 2025
-
IncEventGS: Pose-Free Gaussian Splatting from a Single Event CameraIn Computer Vision and Pattern Recognition (CVPR) 2025Highlights paper, (acceptance ratio: 3.1%)
-
Convex Relaxation for Robust Vanishing Point Estimation in Manhattan WorldIn Computer Vision and Pattern Recognition (CVPR) 2025Best Paper Finalist, Oral, 15 out of 13008 submissions
-
MVControl: Controllable Text-to-3D Generation via Surface-Aligned Gaussian SplattingIn International Conference on 3D Vision (3DV) 2025
-
DreamMesh4D: Video-to-4D Generation with Sparse-Controlled Gaussian-Mesh Hybrid RepresentationIn Neural Information Processing Systems (NeurIPS) 2024
-
BAD-Gaussians: Bundle Adjusted Deblur Gaussian SplattingIn European Conference on Computer Vision (ECCV) 2024
-
BeNeRF: Neural Radiance Fields from a Single Blurry Image and Event StreamIn European Conference on Computer Vision (ECCV) 2024
-
GlobalPointer: Large Scale Plane Adjustment with Bi-Convex RelaxationIn European Conference on Computer Vision (ECCV) 2024
-
RSL-BA: Rolling Shutter Line Bundle AdjustmentIn European Conference on Computer Vision (ECCV) 2024
-
Motion and Structure from Event-based Normal FlowIn European Conference on Computer Vision (ECCV) 2024
-
Event-Aided Time-To-Collision Estimation for Autonomous DrivingIn European Conference on Computer Vision (ECCV) 2024
-
Keypoint-Guided Efficient Pose Estimation and Domain Adaptation for Micro Aerial VehiclesIEEE Transactions on Robotics (T-RO) 2024
-
SCINeRF: Neural Radiance Fields from a Snapshot Compressive ImageIn Computer Vision and Pattern Recognition (CVPR) 2024Highlights paper (acceptance ratio: 2.8%)
-
USB-NeRF: Unrolling Shutter Bundle Adjusted Neural Radiance FieldsIn International Conference on Learning Representations (ICLR) 2024
-
BAD-NeRF: Bundle Adjusted Deblur Neural Radiance FieldsIn Computer Vision and Pattern Recognition (CVPR) 2023
-
MBA-VO: Motion Blur Aware Visual OdometryIn International Conference on Computer Vision (ICCV) 2021Oral presentation (acceptance ratio: 3%)
-
Deep Shutter Unrolling NetworkIn Computer Vision and Pattern Recognition (CVPR) 2020
-
Self-Supervised Linear Motion DeblurringIn Robotics and Automation Letters (RA-L) 2020
-
Project AutoVision: Localization and 3D Scene Perception for an Autonomous Vehicle with a Multi-Camera SystemIn International Conference on Robotics and Automation (ICRA) 2019
-
Efficient 2D-3D Matching for Multi-Camera Visual LocalizationIn International Conference on Robotics and Automation (ICRA) 2019
-
Towards Robust Visual Odometry with a Multi-Camera SystemIn International Conference on Intelligent Robots and Systems (IROS) 2018
-
Robust Dense Mapping for Large-Scale Dynamic EnvironmentsIn International Conference on Robotics and Automation (ICRA) 2018
-
Direct Visual Odometry for a Fisheye-Stereo CameraIn International Conference on Intelligent Robots and Systems (IROS) 2017