search

mengshu1996 / SVBRDF-GAN

20 stars 4 forks source link

why do you group normal and roughness, in other word, why you group diffuse and specular map? #1

Open XueZ-phd opened 4 years ago

XueZ-phd commented 4 years ago

Yezi, hi! I just watch your live on EGSR2020, you have a good presentation, and congraduate for publishing! However, I have one question, as topic，why you group normal and roughness maps? As we see, under the cook-torrance rendering model, the normal map also accounts for diffuse rendering. Looking forward to your response! xue

mengshu1996 commented 4 years ago

Hi！ Thanks a lot for your question : ) At the beginning of our experiments, we find the diffuse map and normal map are highly affected by each other when just using a single encoder to produce 4 maps. Specifically, lightings are baked on the diffuse map and the normal map looks flat with few bumps. This is implausible because highlights are more likely to be caused by bumps and specular. So we use 2 decoders to take the diffuse map and normal map apart. As for roughness and specular, we didn't find many differences between different combinations (RoughnessDiffuse_SpecularNormal or SD_RN), so we just chose one at will. Hope this helps.