Tong Wei

PhD candidate at VRG, Czech Technical University in Prague. weitong@fel.cvut.cz

my_pic.jpg

I am a PhD candidate working at CTU in Prague, supervised by Dr. Dániel Baráth and Prof. Jiří Matas. Before that, I completed my Master’s in Computer Science (AI specialization) at ELTE in Budapest, supervised by Dániel Baráth and Levente Hajder.
My research mainly focuses on subproblems of SfM, including differentiating RANSAC for pose estimation and rigid transformation, as well as improving image retrieval targeted at a good and robust initial pose graph and reconstruction. I am open to internship this year and postdoc/researcher positions starting next year.

news

Feb 21, 2026 Our paper titled “Global-Aware Edge Prioritization for Pose Graph Initialization” has been accepted to CVPR 2026! Big thanks to my amazing co-authors & supervisors: Giorgos Tolias, Jiri Matas, Daniel Barath. :smile: :tada: See you in Denver!
Oct 28, 2024 Our paper called “Breaking the Frame: Visual Place Recognition by Overlap Prediction” has been accepted to WACV 2025. Chat soon in Tucson! :sparkler: :cactus:
Oct 01, 2024 I am working as an academic guest researcher at CVG, ETH Zurich for two months, thanks to Dr. Daniel Barath and Prof. Marc Pollefeys.
Jul 13, 2023 Two first-authored papers has been accepted to ICCV 2023! “Generalized Differentiable RANSAC” and “Adaptive Reordering Sampler with Neurally Guided MAGSAC” Big thanks to our co-authors & supervisors: Yash Patel, Jiri Matas, Daniel Barath, and Sasha Shekhovtsov. :smile:

selected publications

  1. global.png
    Global-Aware Edge Prioritization for Pose Graph Initialization
    Tong Wei, Tolias Giorgos, Jiřı́ Matas, and Daniel Barath
    In CVPR, 2026
  2. vop.png
    Breaking the Frame: Visual Place Recognition by Overlap Prediction
    Tong Wei, Philipp Lindenberger, Jiřı́ Matas, and Daniel Barath
    In WACV, 2025
  3. ransac.png
    Generalized Differentiable RANSAC
    Tong Wei, Yash Patel, Alexander Shekhovtsov, Jiri Matas, and Daniel Barath
    In ICCV, 2023
  4. ar_sampler.png
    Adaptive Reordering Sampler with Neurally Guided MAGSAC
    Tong Wei, Jiri Matas, and Daniel Barath
    In ICCV, 2023