Global optimization and learning for lighting design | TU Wien

Information

Publication Type: Bachelor Thesis
Workgroup(s)/Project(s):
- Rendering and Modeling
Date: September 2024
Date (Start): 15. March 2024
Date (End): 30. September 2024
Matrikelnummer: e11924496
First Supervisor:
- David Hahn
- Michael Wimmer

Abstract

Global optimisation and learning for lighting design is a work in extension of the Tamashii rendering framework by the Rendering and Modeling Group at TU Wien. Tamashii offers a user interface and implementation to perform a local optimisation task on a scene to find the position of the scene’s light object in order to recreate a given target illumination. We extend the existing framework by implementing a global search for the optimum and additional surrogate models to provide machine learning alternatives to apply various optimisation algorithms. We examine the performance of several different optimisation methods with respect to efficiency, accuracy, and versatility. Furthermore, we compare the use of algorithms in combination with surrogate models versus optimising on the Tamashii model directly.

Additional Files and Images

BSc Thesis

Weblinks

No further information available.

BibTeX

@bachelorsthesis{Petersen_Viktoria-2024-SBO,
  title =      "Global optimization and learning for lighting design",
  author =     "Viktoria Petersen",
  year =       "2024",
  abstract =   "Global optimisation and learning for lighting design is a
               work in extension of the Tamashii rendering framework by the
               Rendering and Modeling Group at TU Wien. Tamashii offers a
               user interface and implementation to perform a local
               optimisation task on a scene to find the position of the
               scene’s light object in order to recreate a given target
               illumination. We extend the existing framework by
               implementing a global search for the optimum and additional
               surrogate models to provide machine learning alternatives to
               apply various optimisation algorithms. We examine the
               performance of several different optimisation methods with
               respect to efficiency, accuracy, and versatility.
               Furthermore, we compare the use of algorithms in combination
               with surrogate models versus optimising on the Tamashii
               model directly.",
  month =      sep,
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Research Unit of Computer Graphics, Institute of Visual
               Computing and Human-Centered Technology, Faculty of
               Informatics, TU Wien ",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2024/Petersen_Viktoria-2024-SBO/",
}