Computer graphics researcher
Martin Bálint
I develop machine-learning methods for efficient Monte Carlo rendering and video reconstruction, focusing on adaptive sampling, neural reconstruction, and stochastic optimization.
Doctoral researcher
Max Planck Institute for Informatics
Selected projects
All projects
ACM Transactions on Graphics · SIGGRAPH 2026
Forget Superresolution, Sample Adaptively (when Path Tracing)
An end-to-end pipeline that learns where path-tracing samples matter, then reconstructs a detailed image from the sparse result.
SIGGRAPH 2025 Conference Proceedings
Histogram Stratification for Spatio-Temporal Reservoir Sampling
A stratification method for improving spatiotemporal reservoir sampling.
SIGGRAPH Asia 2023 Conference Proceedings
Joint Sampling and Optimisation for Inverse Rendering
A method that treats sample allocation and stochastic optimization as one coupled inverse-rendering problem.
SIGGRAPH 2023 Conference Proceedings
Neural Partitioning Pyramids for Denoising Monte Carlo Renderings
A learned pyramidal filter that reconstructs low-sample-count path-traced images while avoiding ringing and scale-composition artifacts.
Awards
2023
Future of Graphics and Media Research Grant
Saarland/Intel
Jointly awarded with Karol Myszkowski for a proposal building on Neural Partitioning Pyramids for Denoising Monte Carlo Renderings.
2021
IMPRS-TRUST
Direct doctorate program
2019
Best Paper Award
CESCG
Awarded for Closed Form Transmittance in Heterogeneous Media Using Cosine Noise.
Positions
Jul 2020 – Sep 2020
Research Intern
Max Planck Institute for InformaticsSaarbrücken, Germany
Advisor: Karol Myszkowski
Contributed to Perceptual Model for Adaptive Local Shading and Refresh Rate, published at SIGGRAPH Asia 2021.
Jun 2019 – Aug 2019
Research Intern
University of CambridgeCambridge, United Kingdom
Advisor: Rafał Mantiuk
Developed camera-feedback calibration and banding-mitigation methods for novel displays.
Education
Sep 2021 – Nov 2026 (expected)
Joint Master’s and Ph.D.
Max Planck Institute for InformaticsSaarbrücken, Germany
Advisors: Karol Myszkowski and Hans-Peter Seidel
Thesis: Efficient Monte Carlo Rendering through Learned Sampling and Reconstruction
Sep 2018 – Jun 2021
Computer Science BA, Churchill College
University of CambridgeCambridge, United Kingdom
Thesis: Towards Predictable Transactional Memory
Sep 2011 – Jun 2018
Specialized Mathematics Class
Mihály Fazekas High SchoolBudapest, Hungary
Second place, Microsoft Imagine Cup 2017; first place, Neumann International Software Competition 2017.
Academic service
2026
IPC Member and Session Chair
High-Performance Graphics
Reviewer
SIGGRAPH, SIGGRAPH Asia, Eurographics, Computer Graphics Forum, and Journal of Computer Graphics Techniques