Für eine korrekte Darstellung dieser Seite benötigen Sie einen XHTML-standardkonformen Browser, der die Darstellung von CSS-Dateien zulässt.

You are here

  • English
  • Deutsch

MA

Modern Material Models for Monte-Carlo Path Tracing

Full Thesis Title: 
Modern Material Models for Monte-Carlo Path Tracing
Supervisor: 
Martin Lambers
Status: 
Open
Requirements: 
Computer Graphics III
Short Description: 
In this thesis, the candidate will review the relevant literature both on modern material models and on importance sampling of materials for Monte-Carlo path tracing, and construct a physically plausible material model that is able to reproduce the behavior of a wide variety of real-world materials while still supporting proper importance sampling as far as possible.
Image: 
Thesis Type: 

Anisotropic Smoothed Particle Hydrodynamics

Full Thesis Title: 
Anisotropic Smoothed Particle Hydrodynamics
Supervisor: 
M. Sc. Rene Winchenbach
Status: 
Open
Requirements: 
Computer Graphics IV
Short Description: 
The goal of this method is to evaluate possible benefits that anisotropy can bring to SPH. The student would have to implement such a method, e.g. arxiv.org/abs/astro-ph/9512078 , in our existing simulation framework using CUDA and C++. Due to the difficulty of the method itself, prior knowledge in SPH is definetly beneficial.
Thesis Type: 

Blur kernel estimation by deep learning

Full Thesis Title: 
Blur kernel estimation by deep learning
Supervisor: 
Tak Ming Wong
Status: 
Open
Short Description: 
Recently, computer vision community developed several motion deblurring and single image super-resolution methods, which model the deblurring process by deep neural network. However, estimating an explicit blur kernel still plays an important role in image processing perspective. The aim of this work is to study in blur kernel estimation on conventional RGB image using state-of-arts deep learning technique.
Thesis Type: 
Subscribe to MA