I am trying the streetsurf using waymo datasets. It seems that they all use lotd for encoding, in the code, I cannot find the link between the lotd and NGP. As I read the readmes and the paper, I assumes the lotd can create a cuboid NGP, instead of cubic. Am I right?
Do you build you own ngp like feature encoding structure? is any document explains more details of lotd.
I am trying the streetsurf using waymo datasets. It seems that they all use lotd for encoding, in the code, I cannot find the link between the lotd and NGP. As I read the readmes and the paper, I assumes the lotd can create a cuboid NGP, instead of cubic. Am I right?
Do you build you own ngp like feature encoding structure? is any document explains more details of lotd.