Presentation

<
A Generalizable Light Transport 3D Embedding for Global Illumination
SessionLight Transport
DescriptionWe introduce a scalable 3D embedding that learns global illumination directly from 3D scene configurations. This design scales to environments with millions of triangles, enabling the first generalizable GI learning on complex, high-fidelity indoor scenes, far beyond prior limits. It bypasses per-scene retraining and screen-space artifacts to deliver realistic rendering.
Authors
Bing Xu
University of California San Diego
NVIDIA
Mukund Varma T
University of California San Diego
Cheng Wang
University of California San Diego
Tzu-mao Li
University of California San Diego
Lifan Wu
NVIDIA
Bartlomiej Wronski
NVIDIA
Ravi Ramamoorthi
University of California San Diego
Marco Salvi
NVIDIA
Event Type
Technical Paper
TimeTuesday, 21 July 20262:10pm - 2:20pm PDT
LocationRoom 408 A
Digital Library PDF
Session Time & Location
Sunday, 19 July 20266:00pm - 8:45pm PDTHall K
Tuesday, 21 July 20262:00pm - 3:30pm PDTRoom 408 A
Keywords
Rendering
Registration Categories
Full Conference Supporter
Full Conference
Next PresentationNext PresentationRao-Blackwellized Markov chain Monte Carlo Light Transport
Similar Presentations
Emerging Technologies Demo - A Generalizable Light Transport 3D Embedding for Global Illumination
8DNA: 8D Neural Asset Light Transport by Distribution Learning
EAG-PT: Emission-Aware Gaussians and Path Tracing for Diffuse Indoor Scene Reconstruction and Editing