Mip-NeRF#
A Multiscale Representation for Anti-Aliasing Neural Radiance Fields
Running Model#
ns-train mipnerf
Overview#


The primary modification in MipNeRF is in the encoding for the field representation. With the modification the same mip-NeRF field can be use for the coarse and fine steps of the rendering hierarchy.


In the field, the Positional Encoding (PE) is replaced with an Integrated Positional Encoding (IPE) that takes into account the size of the sample.