Inverse Rendering for Shape, Light, and Material Decomposition using Multi-Bounce Path Tracing and Reservoir Sampling

①Comparison on Albedo of hotdog

The results of Neural-PBIR were extracted directly from their paper since they have not released their code

②Comparison on ours and MIS with sample count 1

MIS combines the two samples obtained from BRDF sampling and light sampling using MIS weight. This approach differs from our method, which directly sample from the product of the two.

Inverse Rendering for Shape, Light, and Material Decomposition using Multi-Bounce Path Tracing and Reservoir Sampling

-

①Comparison on Albedo of hotdog

②Comparison on ours and MIS with sample count 1

③Comparison on w/ and w/o Spatiotemporal reuse