search

chungyiweng / humannerf

HumanNeRF turns a monocular video of moving people into a 360 free-viewpoint video.
MIT License
786 stars 86 forks source link

Eq.16 in supplementary material #69

Closed xyIsHere closed 1 year ago

xyIsHere commented 1 year ago

Dear author,

Thank you for your great work. I'm new to the human nerf area. Could you explain why the equation is defined like this to compute the observation-to-canonical transformation? Is there any reference I can read to understand this equation?

theAdityaRamesh commented 1 year ago

You can refer this paper Vid2Actor (Eq 10,11). HumanNeRF derives many of its parts from this paper.

xyIsHere commented 1 year ago

Thanks a lot! I also checked that paper before but still didn't fully get understand it. Do u know what does the inverse do in this equation? image