Neural Product Importance Sampling via Warp Composition
SessionSampling and Light Transport
DescriptionAchieving high efficiency in modern photorealistic rendering methods hinges on using Monte Carlo sampling distributions that closely approximate the illumination integral estimated for every pixel. Samples are typically generated from a set of simple distributions, each targeting a different factor in the integrand, which are combined via multiple importance sampling. The resulting mixture distribution can be far from the actual product of all factors, leading to sub-optimal variance even for direct-illumination estimation. We present a learning-based method to efficiently importance sample illumination product integrals (e.g., the product of environment lighting and material terms) using normalizing flows. Our neural product sampler composes a flow head warp with an emitter tail warp. The small conditional head is represented by a neural spline flow, while the large unconditional tail is discretized per environment map and its evaluation is instant. If the conditioning is low-dimensional, the head warp can be discretized for even better performance. We demonstrate variance reduction over prior methods on a range of applications comprising complex geometry, materials and illumination.
Event Type
Technical Papers
TimeThursday, 5 December 20241:34pm - 1:46pm JST
LocationHall B5 (2), B Block, Level 5
