Recently, I have been focusing on real-time light transport algorithms, specifically with very constrained (ray) budgets but also investigating the statistical theory. In this context, I have largely published randomized Monte Carlo-style algorithms, but I have also thought about and investigated the applicability of neural networks. Real-time rendering algorithms require denoisers, so I have been closely involved with both hand-crafted and neural denoisers (and, similar, supersampling algorithms). I also think about and understand state-of-the-art accelerations structures (BVHs and the like, for rays, photons, [ir-]radiance, lights, etc.).

I also actively follow real-time rasterization-based rendering algorithms (and published widely on that topic up until 5 years ago). I have up-to-date knowledge of graphics hardware, and interact frequently with game developers. I view real-time and offline rendering to be converging, so I'm happy to review any paper that matches my experience but is not "real-time."

I am reasonably knowledgeable on hyperscale/distributed/cloud rendering, physically based materials, foveated rendering / eye tracking, VR displays, and neural content generation (geometry, materials, faces), though I do not actively publish in those areas.