Giters

TRI-ML / dd3d

Official PyTorch implementation of DD3D: Is Pseudo-Lidar needed for Monocular 3D Object detection? (ICCV 2021), Dennis Park*, Rares Ambrus*, Vitor Guizilini, Jie Li, and Adrien Gaidon.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

The "remapped" for the pretrained model ?

feifei-Liu opened this issue · comments

commented

Hello, what is the "remapped" mean in the pretrained model (https://tri-ml-public.s3.amazonaws.com/github/dd3d/pretrained/depth_pretrained_dla34-y1urdmir-20210422_165446-model_final-remapped.pth)? If I want to train DD3D using DLA34 (only pretrained on Imagenet), what transformation need to do for DLA model parameters ?

commented

Hello @feifei-Liu , remapped only means the keys in the state_dict of the checkpoint is renamed. The checkpoints are from my old experiments, and I needed to rename the keys to comply with the current version of code.