Making the next generation of virtual reality graphics possible.
Virtual reality and Graphics Architectures (VGA) group
Tell Me More

About

Welcome to the home page of the Virtual reality and Graphics Architectures (VGA) research group in Tampere University. The VGA group is active in graphics research since 2015, building on long experience of Tampere University of Technology on Customized Parallel Computing platforms. Our research aims to make photorealistic graphics algorithms such as path tracing fast enough for real-time rendering.

Path tracing can produce visually pleasing and realistic frame sequences of 3D worlds. However, it is computationally demanding and therefore real-time path tracing is currently limited to powerful desktop computers. Our research aims to enable photorealistic rendering in mobile devices with restrictive power and energy budgets by exploring new graphics hardware architectures and software systems.


News

February 28, 2024: Hierarchical Bitmask Implicit Grids codes released
Our reference implementation for "Hierarchical Bitmask Implicit Grids for Efficient Point-In-Volume Queries on the GPU" is now available at https://github.com/vga-group/hierarchical-bitmask-implicit-grids (GPLv3 license), with instructions on how to reproduce our benchmark results.
January 29, 2024: New publications added
  • Erwan Leria, Markku Mäkitalo, Pekka Jääskeläinen:
    Real-Time Stereoscopic Image-Parallel Path Tracing (preprint)
  • Julius Ikkala, Tuomas Lauttia, Pekka Jääskeläinen, Markku Mäkitalo:
    Hierarchical Bitmask Implicit Grids for Efficient Point-In-Volume Queries on the GPU (preprint)
October 27, 2023: A master's thesis added
  • Tuomas Lauttia:
    Contemporary Specular Denoising Algorithms in Real-Time Path Tracing (link)
In this work, two contemporary denoising algorithms, Spatiotemporal Variance-Guided Filtering (SVGF) and Blockwise Multi-Order Feature Regression (BMFR) were reimplemented into Tauray and evaluated in terms of visual quality and execution time using modern hardware with ray tracing support.
May 8, 2023: TauBench v1.1 released
Version 1.1 of our TauBench dataset is now available. We made the 3D scene files much more compact (about 900 MB per scene, instead of 2.7 GB per scene) for significantly faster loading times and improved usability, and fixed various minor bugs. Download the updated dataset and see the technical notes of the v1.1 release, including temporal PSNR benchmark results for the TAA and SVGF algorithms.
January 18, 2023: BMFR denoising pipeline extended by AMD
Researchers at the Advanced Rendering Research Group at AMD have published a regression-based path tracing denoiser based on our BMFR pipeline. Their method is called WALR (Weighted À-Trous Linear Regression), and it extends BMFR by replacing our blockwise regression stage with an edge-aware per-pixel denoiser stage. They use the Cholesky decomposition for solving the weighted regression problem, as opposed to the QR decomposition used in BMFR. They report execution times similar to BMFR at an improved quality, with the caveat that they assume only diffuse indirect lighting. Their paper is available on GPUOpen.
October 11, 2022: Tauray rendering engine released
We have just released our open-source, cross-platform real-time 3D renderer called Tauray. The primary focus of Tauray is in using photorealistic path tracing techniques to generate real-time content for multi-view displays, such as VR headsets and light field displays. Ray tracing hardware acceleration as well as multi-GPU rendering is supported. You can find the source code and more information on the Tauray GitHub page. The paper preprint is also available, including benchmarks against other real-time path tracers. We will present the paper at SIGGRAPH Asia Technical Communications on December 8, 2022.
September 14, 2022: New publications added
  • Markku Mäkitalo, Erwan Leria, Julius Ikkala, Pekka Jääskeläinen:
    Real-Time Light Field Path Tracing (preprint)
  • Erwan Leria, Markku Mäkitalo, Julius Ikkala, Pekka Jääskeläinen:
    Dynamic Load Balancing for Real-Time Multiview Path Tracing on Multi-GPU Architectures (preprint)
February 8, 2022: TauBench v1.0 dataset and new publications added
  • TauBench v1.0 is now released. It is a permissively licensed (CC-BY-NC-SA 4.0) dynamic benchmark for graphics rendering, especially for benchmarking temporal reuse methods. Download the dataset and see more info here.
  • Joel Alanko, Markku Mäkitalo, Pekka Jääskeläinen:
    TauBench: Dynamic Benchmark for Graphics Rendering (conference paper preprint)
  • Petrus Kivi, Markku Mäkitalo, Jakub Žádník, Julius Ikkala, Vinod Kumar Malamal Vadakital, Pekka Jääskeläinen:
    Real-Time Rendering of Point Clouds With Photorealistic Effects: A Survey (link)
September 16, 2021: New publication added
  • Julius Ikkala, Petrus Kivi, Joel Alanko, Markku Mäkitalo, Pekka Jääskeläinen:
    DDISH-GI: Dynamic Distributed Spherical Harmonics Global Illumination (link)

Datasets

Example rendered frame of TauBench dataset

TauBench: Dynamic Benchmark for Graphics Rendering

2022, Alanko, J., Mäkitalo, M., and Jääskeläinen, P., In Proceedings of the 17th International Conference on Computer Graphics Theory and Applications (GRAPP).

Publications

Journal Articles

Dynamic load balancing of multiple viewpoints on multiple GPUs

Dynamic Load Balancing for Real-Time Multiview Path Tracing on Multi-GPU Architectures

2022, Leria, E., Mäkitalo, M., Ikkala, J. and Jääskeläinen, P., In Virtual Reality & Intelligent Hardware (VRIH).

Splat-based representation for level-of-detail point cloud rendering

Real-Time Rendering of Point Clouds With Photorealistic Effects: A Survey

2022, Kivi, P., Mäkitalo, M., Žádník, J., Ikkala, J., Malamal Vadakital, V. K. and Jääskeläinen, P., In IEEE Access, volume 10.

Framewise reprojection quality graph

Systematic Evaluation of the Quality Benefits of Spatiotemporal Sample Reprojection in Real-Time Stereoscopic Path Tracing

2020, Mäkitalo, M., Kivi, P., and Jääskeläinen, P., In IEEE Access, Volume 8.

Noisy input versus BMFR result comparison on Sponza

Blockwise Multi-Order Feature Regression for Real-Time Path Tracing Reconstruction

2019, Koskela M., Immonen K., Mäkitalo M., Foi A., Viitanen T., Jääskeläinen P., Kultala H., and Takala J., In ACM Transactions on Graphics (TOG), Volume 38 Issue 5.

Part of the PLOCTree sweep pipeline

PLOCTree: A Fast, High-Quality Hardware BVH Builder

2018, Viitanen, T., Koskela, M., Jääskeläinen, P., Tervo, A., and Takala, J., In Proceedings of the ACM on Computer Graphics and Interactive Techniques 1, 2, Article 35 (August 2018).

Instantaneous foveated preview: noisy Sponza frame

Instantaneous foveated preview for progressive Monte Carlo rendering

2018, Koskela, M., Immonen, K., Viitanen, T., Jääskeläinen, P., Multanen, J., and Takala, J., In Springer Computational Visual Media, 4th of April 2018.

Part of the MergeTree pipeline

MergeTree: A Fast Hardware HLBVH Constructor for Animated Ray Tracing

2017, Viitanen, T., Koskela, M., Jääskeläinen, P., Kultala, H., and Takala, J., In ACM Transactions on Graphics (TOG), Volume 36 Issue 5.

Quantized BVH: example scene (cloth) with a large amount of geometry with a degenerate axis

Fast Hardware Construction and Refitting of Quantized Bounding Volume Hierarchies

2017, Viitanen, T., Koskela, M., Jääskeläinen, P., Immonen, K. and Takala, J., In Computer Graphics Forum, Volume 36 Issue 4.


Conference Proceedings

Sponza with a lot of colourful lights, rendered using the HBIG data structure

Hierarchical Bitmask Implicit Grids for Efficient Point-In-Volume Queries on the GPU

2024, Ikkala, J., Lauttia, T., Jääskeläinen, P. and Mäkitalo, M., In Proceedings of the 19th International Conference on Computer Graphics Theory and Applications (GRAPP).

Bistro exterior, showing different intermediate results (reprojection and per-GPU missing pixels, hole-filling, denoising)

Real-Time Stereoscopic Image-Parallel Path Tracing

2024, Leria, E., Mäkitalo, M. and Jääskeläinen, P., In Engineering Reality of Virtual Reality, Electronic Imaging 2024.

Logo of the Tauray rendering engine

Tauray: A Scalable Real-Time Open-Source Path Tracer for Stereo and Light Field Displays

2022, Ikkala, J., Mäkitalo, M., Lauttia, T., Leria, E. and Jääskeläinen, P., ACM SIGGRAPH Asia Technical Communications.

Real-time light field path tracing

Real-Time Light Field Path Tracing

2022, Mäkitalo, M., Leria, E., Ikkala, J. and Jääskeläinen, P., In Proceedings of the 39th Computer Graphics International (CGI).

Example rendered frame of TauBench dataset

TauBench: Dynamic Benchmark for Graphics Rendering

2022, Alanko, J., Mäkitalo, M., and Jääskeläinen, P., In Proceedings of the 17th International Conference on Computer Graphics Theory and Applications (GRAPP).

Three example scenes rendered with DDISH-GI. One has glossy reflections, one emissive materials, and one diffuse indirect lighting

DDISH-GI: Dynamic Distributed Spherical Harmonics Global Illumination

2021, Ikkala, J., Kivi, P., Alanko, J., Mäkitalo, M. and Jääskeläinen, P., In Proceedings of the 38th Computer Graphics International (CGI).

CNN pipeline for foveated path tracing reconstruction

Machine Learning is the Solution Also for Foveated Path Tracing Reconstruction

2020, Lotvonen, A., Koskela, M., and Jääskeläinen, P., In Proceedings of the 15th International Conference on Computer Graphics Theory and Applications (GRAPP).

Path traced and denoised frame with the Stanford Bunny in a room with mirrors, transformed from visual-polar space into Cartesian screen space

Foveated Real-Time Path Tracing in Visual-Polar Space

2019, Koskela, M., Lotvonen, A., Mäkitalo, M., Kivi, P., and Jääskeläinen, P., In Proceedings of the Eurographics Symposium on Rendering.

Noisy Living Room frame, highlighting pixels where stereo reprojection from the other eye did not succeed

Reducing Computational Complexity of Real-Time Stereoscopic Ray Tracing with Spatiotemporal Sample Reprojection

2019, Mäkitalo, M., Kivi, P., Koskela, M., and Jääskeläinen, P., In Proceedings of the 14th International Conference on Computer Graphics Theory and Applications (GRAPP).

Sports car with shiny areas that are potentially difficult for real-time path tracing

Sparse Sampling for Real-Time Ray Tracing

2018, Viitanen, T., Koskela, M., Immonen, K., Mäkitalo, M., Jääskeläinen, P. and Takala, J., In Proceedings of the 13th International Conference on Computer Graphics Theory and Applications (GRAPP).

Foveated instant preview: noisy Sponza frame

Foveated Instant Preview for Progressive Rendering

2017, Koskela, M., Immonen, K., Viitanen, T., Jääskeläinen, P., Multanen, J. and Takala, J., In ACM SIGGRAPH Asia Technical Briefs.

Path traced Cornell Box with a reflective sphere and a refractive Stanford Bunny, caustics, and soft shadows

Foveated Path Tracing: A Literature Review and a Performance Gain Analysis

2016, Koskela, M., Viitanen, T., Jääskeläinen, P. and Takala, J., In Proceedings of the 12th International Symposium on Visual Computing (ISVC).

Hairball test scene

Multi Bounding Volume Hierarchies for Ray Tracing Pipelines

2016, Viitanen, T., Koskela, M., Jääskeläinen, P. and Takala, J., In ACM SIGGRAPH Asia Technical Briefs.

Shiny Utah Teapot inside the Sponza scene

Half-Precision Floating-Point Ray Traversal

2016, Koskela, M., Viitanen, T., Jääskeläinen, P. and Takala, J., In Proceedings of the 11th International Conference on Computer Graphics Theory and Applications (GRAPP).

Dragon, partially rendered and partially showing the constructed BVH structure

MergeTree: A HLBVH Constructor for Mobile Systems

2015, Viitanen, T., Koskela, M., Jääskeläinen, P., Kultala, H., and Takala, J., In ACM SIGGRAPH Asia Technical Briefs.

The Buddha's hand, showing the workload differences between the half-precision BVH and the hierarchical half-precision BVH

Using Half-Precision Floating-Point Numbers for Storing Bounding Volume Hierarchies

2015, Koskela, M., Viitanen, T., Jääskeläinen, P., Takala, J., and Cameron, K., In Proceedings of the 32nd Computer Graphics International Conference (CGI).

TTA processor architecture example

Programmable and Scalable Architecture for Graphics Processing Units

2009, Lama, C., Jääskeläinen, P., and Takala, J., In Proceedings of the International Conference on Embedded Computer Systems: Architectures, MOdeling and Simulation (SAMOS).


Theses

Graph showing the execution time breakdown for different components of the BMFR denoiser on the Sponza scene

Contemporary Specular Denoising Algorithms in Real-Time Path Tracing

2023, Lauttia, T., Master of Science thesis, Tampere University.

Neural network performance comparison graph

Fast Convolutional Neural Networks for Real-Time Path Tracing Denoising

2021, Lotvonen, A., Master of Science thesis, Tampere University.

Example frame showing an open outdoor area at the end of a brick tunnel

Dynamic Benchmark for Graphics Rendering

2021, Alanko, J., Master of Science thesis, Tampere University.

Path traced outdoor urban scene, highlighting pixels where reprojection did not succeed

Sample Reprojection for Real-Time Light Field Path Tracing

2021, Laukkanen, L., Master of Science thesis, Tampere University.

Utah Teapot in a Cornell Box, with a grid of spherical light probes visualized

Real-Time Path Traced Spherical Harmonics Probes

2020, Ikkala, J., Master of Science thesis, Tampere University.

Path traced Cornell Box with a reflective sphere, refractive Dragon, and soft shadows

Foveated Path Tracing with Fast Reconstruction and Efficient Sample Distribution

2020, Koskela, M., Doctoral dissertation, Tampere University.

A polynomial surface intersected by a ray in 3D space

Ray Tracing Methods for Point Cloud Rendering

2019, Kivi, P., Master of Science thesis, Tampere University.

Classroom scene with the central area in focus and the periphery blurred

Scalable Parallel Path Tracing

2019, Ikkala, J., Bachelor of Science thesis, Tampere University.

Example frame of the animated Toasters scene

Hardware Accelerators for Animated Ray Tracing

2018, Viitanen, T., Doctoral dissertation, Tampere University of Technology.

Noisy versus denoised comparison frame with a foveated sample distribution, showing the front of a car

Real-Time Noise Removal in Foveated Path Tracing

2017, Immonen, K., Master of Science thesis, Tampere University of Technology.

Cornell Box with a reflective sphere, refractive TUT logo, caustics, and soft shadows

Software-Based Ray Tracing for Mobile Devices

2015, Koskela, M., Master of Science thesis, Tampere University of Technology.

Supervised Theses

Title
Author(s)
Augmented Reality Framework and Demonstrator
2018, Rasti, D. & Rasti, A., Master of Science thesis
Simplification of 3D Computer-Aided Design Models to Improve Rendering Performance
2019, Heinonen, A., Master of Science thesis
Interactive Global Illumination for Static Computer-Aided Design Models Using Signed Distance Fields
2021, Kärnä, J., Master of Science thesis

Contact Us

Send email to to contact us.