| Session Name: | Fast Denoising With Self Stabilizing Recurrent Blurs (Presented by NVIDIA) |
| Speaker(s): | Dmitry Zhdan |
| Company Name(s): | NVIDIA |
| Track / Format: | Visual Arts |
Did you know free users get access to 30% of content from the last 2 years?
Get your team full access to the most up to date GDC content
| Overview: | In this topic NVIDIA is going to discuss latest advancements in non-DL based denoising. Basing on previous work from Metro Exodus, a new method has been introduced which is based on recurrent blur too, but it has got a lot of improvements, like: better overall performance, cleaner results, specular denoising support, fast data reconstruction, better bilateral weighting and some others. Besides the algorithm overview the topics will be covered in the talk:n- mipmapping of incoming radiance - is it worth it?n- proper controlling of blur radius to avoid over-blurringn- accurate dis-occlusion detectionn- specular tracking without specular motionn- fast and high amplitude noise free data reconstruction of regions with discarded historyn- exponential versus linear accumulation. Why linear accumulation is better?n- how to compute bilateral weights in spatial passes?n- input signal compression - should it be used or not?n- how to fight with temporal lag |
Recommended
