Camera Geometry Problems in Computer Vision

Website: https://sites.google.com/view/camgeo2024/home
Date: March 18, 2024 (14:20 - 18:00)
Room: TBD

  • Zuzana Kukelova (Czech Technical University (CTU) in Prague)
  • Laura Leal-Taixé (NVIDIA)
  • Daniel Barath (ETH Zürich)
  • Viktor Larsson (Lund University)
  • Tomas Pajdla (Czech Technical University (CTU) in Prague)
  • Torsten Sattler (Czech Technical University (CTU) in Prague)

The aim of this tutorial is to raise awareness of the tools for camera geometry estimation (solvers and solver generators) that are nowadays at the disposal of 3D vision researchers and practitioners and which problems can and cannot be efficiently solved at the moment. To this end, the tutorial will discuss current state-of-the-art minimal and non-minimal solvers, explain how to implement them and how to use them in practice, as well as give examples of their use in applications.

3D Gaussian Splatting

Date: March 18, 2024 (14:30 - 18:00)
Room: TBD

  • Georgios Kopanas (Google)
  • Bernhard Kerbl (TU Wien)
  • Jonathon Luiten (Meta Reality Labs)
  • Antoine Guédon (ENPC)

3D Gaussian Splatting(3DGS) suggests a rasterization-based method for reconstructing radiance fields. This allows for high-quality reconstruction that achieves high speed and performance even for low-end devices. Gaussian Splatting motivated practitioners to build a significant amount of tools and applications. 3DGS has also been integrated with frameworks like Unity, Unreal, three.js, and NeRFStudio. On top of that, a significant amount of research that uses 3DGS has been released, spanning different topics: from dynamic and human reconstruction to SLAM, 3D Generative Models, and many more. In this tutorial, we will first explain 3DGS, we will give practical suggestions about using the different tools we find useful in a research context, and talk about the progress in the research domain.