Active Exploration for Neural Global Illumination of Variable Scenes
Event Type
Technical Paper
Interest Areas
Research & Education
Presentation Types
Registration Categories
Full Conference Supporter
Full Conference
Virtual Conference Supporter
Virtual Conference
Exhibitor Additional Full Conference
Exhibitor Full Conference
DescriptionWe introduce a novel active exploration method using Markov Chain Monte Carlo, which explores the space D of possible training data instances, generating ground truth renderings that best help training. After training, our method allows interactive rendering of hard light transport paths (e.g., complex caustics) and scene manipulation.