Interactive Ray Tracing

In recent years, there has been a renewed interest in real-time ray tracing for interactive applications. This is due to many factors: firstly, processor speed has continued to rise at exponential rates as predicted by Moore’s Law and is approaching the raw computational power needed for interactive ray tracing. Secondly, ray tracing algorithms can be highly parallelized on shared memory and distributed memory systems. Therefore, the current hardware trend towards desktop systems with multi-core CPUs and programmable GPUs can be used to accelerate ray tracing. Finally, recent algorithmic improvements that exploit ray coherence can achieve a significant improvement in rendering time.

We have active research projects in the area of interactive ray tracing dealing with problems such as ray tracing massive datasets and deformable models.

• Dynamic Scene Benchmarks, Datasets
• SATO: Surface-Area Traversal Order for Shadow Ray Tracing, CGF 2014
• RayCore: A ray-tracing hardware architecture for mobile devices, ACM TOG 2014
• HART:A Hybrid Architecture for Ray Tracing Animated Scenes, IEEE TVCG 2014
• Selective Ray Tracing for Interactive High-quality Shadows, Tec. Report 2009
• Fast BVH Construction on GPUs, CGF 2009
• ReduceM: Interactive and Memory Efficient Ray Tracing of Large Models, EGSR 2008
• AD-Frustum: Adaptive Frustum Tracing for Interactive Sound Propagation, IEEE Visualization 2008
• Ray-strips: A Compact Mesh Representation for Interactive Ray Tracing, RT 2007
• Ray Tracing Dynamic Scenes Using Selective Restructuring, EGSR 2007
• Interactive Sound Propagation in Dynamic Scenes Using Frustum Tracing, IEEE Visualization 2007
• R-LODs: Fast LOD-Based Ray Tracing of Massive Models, PG 2006
• RT-DEFORM: Interactive Ray Tracing of Dynamic Scenes Using BVHs, RT 2006
• Cache-Efficient Layouts of Bounding Volume Hierarchies, EUROGRAPHICS 2006