Miša Korać

Zdravo! I am a PhD candidate at Computer Graphics Lab in Saarland University, advised by Prof. Dr.-Ing. Philipp Slusallek, currently finishing my thesis. My research revolves around physically based light transport simulation, which aims to create synthetic images indistinguishable from real ones. More precisely, my PhD work focuses on how correlation properties can be harnessed throughout the Monte Carlo rendering pipeline. Building on this, I am interested in continuing my research at the intersection of neural methods and light transport.
During my PhD, I was awarded a Marie Skłodowska-Curie fellowship as part of the PRIME ITN at the German Research Center for Artificial Intelligence (DFKI), which continued to support me afterwards, along with write-up funding from Saarland University. I also interned as a researcher at Maxon, the makers of Cinema 4D and the Redshift renderer. Before coming to Saarbrücken as a fast-track PhD student, I completed my 4-year Bachelor’s studies at the School of Electrical Engineering, University of Belgrade, during which I had the wonderful opportunity to work with doc. Ing. Jaroslav Křivánek.
Besides graphics, I enjoy building my own synthesizer and have also done some research in acoustics, a field with many parallels to light transport. I have always been drawn to design, from architecture to cars, and my grandmother, an engineer on the Galeb G-2, passed her fascination with aeronautics on to me.

Research Interests

  • Monte Carlo Denoising
  • Correlated Sampling
  • Importance Sampling (Path Guiding)
  • Perception
  • Neural Representations

Coming soon

Perceptual error optimization for Monte Carlo animation rendering

ACM SIGGRAPH Asia 2023 (Conference Track)
Miša Korać, Corentin Salaün, Iliyan Georgiev, Pascal Grittmann, Philipp Slusallek, Karol Myszkowski and Gurprit Singh

Research Engineer (PRIME Secondment and Internship)

Maxon Computer

Research in the fields of path guiding and correlated sampling.
Supervised by Dr. Bernd Lutz and Mauro van de Vlasakker.

Junior Researcher

German Research Center for Artificial Intelligence (DFKI)

Research on rendering algorithms for training autonomous driving systems as part of PRIME ITN.
Advised by Prof. Dr.-Ing. Philipp Slusallek.

Research Assistant

Computer Graphics Chair, Saarland University

Research in the fields of path guiding and blue-noise sampling.
Advised by Prof. Dr.-Ing. Philipp Slusallek.

G2 Render Engine

Personal Project and B.Sc. Thesis

Physically based unbiased renderer constructed to use both CPU and GPU. Implements path tracing and vertex connection and merging light transport algorithms. Inspired by the Mitsuba 0.6 renderer.

Light Transport Simulation:
The Science Behind Photorealistic Images

Invited talk at the School of Electrical Engineering, University of Belgrade

Blue Noise Sampling:
From Halftoning to Perceptual Animation Rendering

Tech Talk at Maxon Computer

Thesis Supervision

Implementing and evaluating DDGI resampling in a real-time renderer
Eric Windholz Bachelor's Thesis