Dear author, thanks for the amazing work. The paper said that a LookUp Table is precomputed for faster running time of the model. However, I cannot see where the functions in script/view_transform module called in the python code of the model or how the LUT is used during training or testing of the model. Since I want to adapt the FastBEV View transformer to my model for a different BEV perception task (lane detection). Can you give me some hints?
Thank you