About

Yan Xu is a fourth-year PhD candidate in the Multimedia Laboratory, Chinese University of Hong Kong, advised by Prof. Hongsheng Li and Prof. Xiaogang Wang. Before this, he received his M.S. degree in the State Key Lab of CAD&CG, Zhejiang University, supervised by Prof. Guofeng Zhang and Prof. Hujun Bao. During his PhD, he fortunately worked with researchers from research institutes worldwide, including MIT, Stanford, HKU, Reality Labs Research, Baidu Research, etc.

His current research interests revolve around 3D computer vision. His long-term goal is to create an intelligent system that can perceive, understand and interact with the physical world.

Recent News

Jun 27, 2023: One paper is accepted to RA-L!
Oct 19, 2022: Give a talk at Baidu Research!
Sep 10, 2022: One paper is accepted to CoRL 2022!
Aug 4, 2022: One paper is accepted to SIGGRAPH Asia 2022!
Jul 9, 2022: One paper is accepted to TIP 2022!
Jun 20, 2022: Start a research scientist internship in Reality Labs Research!
Mar 3, 2022: Two papers are accepted to CVPR 2022!
Dec 17, 2021: One paper is accepted to RAL 2022!

Selected Projects

  • Robotic Perception: depth completion, scene representation, object pose estimation, concept understanding
  • State Estimation: odometry, localization

CVPR 2022: Visual Grounding
CoRL 2022: Embodied Concept
Learner
CVPR 2022: 6-DoF Object
Pose Estimation

ICCV 2023: Urban-level Neural
Radiance Field

Publications

Urban Radiance Field Representation with Deformable Neural Mesh Primitives.

International Conference on Computer Vision (ICCV), 2023.

PDF PROJECT PAGE

NeRF-Loc: Transformer-Based Object Localization Within Neural Radiance Fields

IEEE Robotics and Automation Letters (RA-L), 2023.

PDF

Embodied Concept Learner: Self-supervised Learning of Concepts and Mapping through Instruction Following

Conference on Robot Learning (CoRL), 2022.

PDF

NeuralMarker: A Framework for Learning General Marker Correspondence

SIGGRAPH Asia, 2022.

PDF

RNNPose: Recurrent 6-DoF Object Pose Refinement with Robust Correspondence Field Estimation and Pose Optimization

Conference on Computer Vision and Pattern Recognition (CVPR), 2022.

PDF CODE

Improving Visual Grounding with Visual-Linguistic Verification and Iterative Reasoning

Conference on Computer Vision and Pattern Recognition (CVPR), 2022.

PDF CODE

IEEE Robotics and Automation Letters (RA-L), 2022 (also accepted by ICRA 2022).

PDF CODE

PDNet: Towards Better One-stage Object Detection with Prediction Decoupling

IEEE Transactions on Image Processing (TIP), 2022.

PDF

VS-Net: Voting with Segmentation for Visual Localization.

Conference on Computer Vision and Pattern Recognition (CVPR), 2021.

PDF CODE
SelfVoxeLO: Self-supervised LiDAR Odometry with Voxel-based Deep Neural Networks.

Conference on Robot Learning (CoRL), 2020.

PDF

SSN: Shape Signature Networks for Multi-class Object Detection from Point Clouds .

European Conference on Computer Vision (ECCV), 2020.

PDF

Depth Completion from Sparse LiDAR Data with Depth-normal Constraints.

International Conference on Computer Vision (ICCV), 2019.

PDF

Prior Guided Dropout for Robust Visual Localization in Dynamic Environments.

International Conference on Computer Vision (ICCV), 2019.

PDF

Teaching Assistant

  • ELEG2401 Introduction to Embedded systems, Fall 2019
  • TECS2461 Machine Intelligence — Principles and Applications, Spring 2020
  • ELEG2401 Introduction to Embedded systems, Fall 2020
  • ELEG1130 Multivariable Calculus for Engineers, Spring 2021

© Yan Xu 2023 · Thanks for your visit.