News:

  • [2021.07] Our paper "MLVSNet: Multi-level Voting Siamese Network for 3D Visual Tracking" is accepted to ICCV 2021. Congrats to Zhoutao!
  • [2021.07] Our paper "VENet: Voting Enhancement Network for 3D Object Detection" is accepted to ICCV 2021. Congrats to Qian!
  • [2021.07] Our paper "Robust and Accurate RGB-D Reconstruction with Line Feature Constraints" is accepted to IROS 2021. Congrats to Yabin!
  • [2021.06] Our paper "Sewer Pipe Defect Detection via Deep Learning with Local and Global Feature Fusion" is accepted to Automation in Construction. Congrats to Dawei!
  • [2021.06] Our paper "Robust and Accurate RGB-D Reconstruction with Line Feature Constraints" is accepted to RAL. Congrats to Yabin!
  • [2021.05] Our paper "Gangue Localization and Volume Measurement" is accepted to TIM. Congrats to Jianping!
  • [2021.04] Our paper "Multi-feature Fusion VoteNet for 3D Object Detection" is accepted to TOMCCAP. Congrats to Zhoutao!
  • [2021.04] Our paper "Part-in-whole Point Cloud Registration for Aircraft Partial Scan Automated Localization" is accepted to CAD. Congrats to Qian!
  • [2021.03] Our paper "Automatic Defect Detection and Segmentation of Tunnel Surface Using Modified Mask R-CNN" is accepted to Measurement. Congrats to Yingying!
  • [2021.03] Our paper "Vote-based 3D Object Detection with Context Modeling and SOB-3DNMS" is also accepted to International Journal of Computer Vision. Congrats to Qian!
  • [2021.02] Our paper "Reinforcement Learning-based Visual Navigation with Information-Theoretic Regularization" is also accepted to ICRA. Congrats to Qiaoyun!
                                                                         more >>

Join Us

If you are interested in digital geometry processing, large-scale 3D data processing, 3D scene analysis and understanding, or 3D measurement and inspection, and would like to join the group as a master or PhD student, postdoc or visiting researcher, please contact Prof. Jun Wang (wjun@nuaa.edu.cn) to discuss opportunities in the group.

About us

We study visual computing the computational principles on 3D data capturing, processing, reconstruction, and high-level 3D scene analysis and understanding. We are interested in building robot-based systems that automatically understand visual scenes, both inferring the semantics and modeling 3D structures.

At the moment, we focus on exploiting 3D geometry data for scene analysis, understanding and reconstruction (e.g. driven by RGB-D data, 3D point clouds, and polygonal mesh models). Additionally, we are also developing techniques on 3D measuring and inspection on industrial products and large-scale construction structures.