You are here
I studied physics at the University of Innsbruck and graduated in Quantum Optics at the Department of Theoretical Physics. One of the main problems of rendering complex scenes is, that no concept or theory exist, which predicts artifacts in rendered pictures. Now, the aim of this project is to develop a perception-based model for detecting artifacts in computer-generated renders, based on perceptual judgments of ‘noisiness’ and other image properties. To reach this ambitious aim, there are several steps to follow.
- First, we have to measure the subjective judgments of noisiness by creating a dataset of computer renderings of complex scenes and and then, perform their evaluation by a set of psychophysical experiments. In this experiment the observers compare how artifacts appears in the images.
- Given the previous results, we will identify the possible set of image properties that could be correlated with the perception of the artifacts produced in the renderings.
- Those properties are the basis for the development of an artifacts predicting model, which will lead us to the development of a method for reducing artifacts or adaptive termination of rendering to avoid artifacts.