Closed NagabhushanSN95 closed 1 year ago
The t_vals in nerf are the distances in the z-axis direction, and t_vals * directions are translated into the actual distance between the two points. And nerfacc uses t_start and t_end to calculate the distance, which is the actual distance between two points.
Thank you for the clarification.
In the original nerf implementation,
directions
andviews
were two different variables. Essentially,views
were unit vectors alongdirections
.views
were used as input to the second NeRF MLP, whereasdirections
were used to compute the location of 3D points aspts_loc = origins + t_vals * directions
. Butrays_marching()
takes only the normalizedviews
as input, which means the same is passed tosigma_fn
and thus 3D point locations are computed aspts_loc = origins + t_vals * views
.Wouldn't this cause a problem? Or did I understand something incorrectly?