Please note: This PhD seminar will take place online.
Wenyou Wang, PhD candidate
David R. Cheriton School of Computer Science
Supervisor: Professor Toshiya Hachisuka
Global illumination (GI) is essential for realistic rendering but remains computationally expensive due to the complexity of simulating indirect light transport. Recent neural methods have mainly relied on per-scene optimization, sometimes extended to handle changes in camera or geometry. Efforts toward cross-scene generalization have largely stayed in 2D screen space, such as neural denoising or G-buffer based GI prediction, which often suffer from view inconsistency and limited spatial understanding.
This work proposes a generalizable 3D light transport embedding that approximates global illumination directly from 3D scene configurations, without using rasterized or path-traced cues. Each scene is represented as a point cloud with geometric and material features. A scalable transformer models global point-to-point interactions to encode these features into neural primitives. At render time, each query point retrieves nearby primitives via nearest-neighbor search and aggregates their latent features through cross-attention to predict the desired rendering quantity. The method demonstrates results on diffuse global illumination prediction across diverse indoor scenes with varying layouts, geometry, and materials. The embedding trained for irradiance estimation can be quickly adapted to new rendering tasks with limited fine-tuning. Preliminary results are also presented for spatial-directional radiance field estimation for glossy materials, showing how the normalized field can accelerate unbiased path guiding. This approach highlights a path toward integrating learned priors into rendering pipelines without explicit ray-traced illumination cues.