Details

Type

  • Bachelor Thesis
  • Student Project
  • Master Thesis

Persons

1-2

Description

Imagine you want to digitize an entire building for a game. Normally, you would need to add all the light sources manually. While position might be ok, light color and direction are already difficult to estimate. Light sources that are only indirectly visible are especially hard. 

 

Light-source estimation example. Red: Areas with high intensity that can be modeled as light sources. Yellow: indirect lighting shows the direction to a light source. Having both geometry and images allows us to estimate the origin of such phenomena. Source: Lumion via X

 

Tasks

Depending on the type of thesis/project, some of these:

  • Estimating light sources in 3D space, not just relative to a photo.
  • Estimating indirectly visible light sources.
  • Estimating different types of light sources.
  • Interdependence of light colors and material colors.

Requirements

  • Knowledge of English language (source code comments and final report should be in English)
  • Knowledge of Python and Numpy
  • Bonus for knowledge of Scipy, Pytorch and CUDA
  • Knowledge of deep-learning basics
  • Bonus for knowledge of Gaussian Splatting and generative models

Environment

The project should be implemented platform independent (Linux, Windows).

A bonus of €500/€1000 if completed to satisfaction within an agreed time-frame of 6/12 months (PR/BA or DA)

Responsible

For more information please contact Philipp Erler, Peter Kán.