About the train/val dataset in Rope3D

ADLab-AutoDrive / BEVHeight

An official code release of our CVPR'23 paper, BEVHeight

MIT License

200 stars 25 forks source link

Open pppp-k opened 6 months ago

pppp-k commented 6 months ago

It seems you use gen_info_rope3d to randomly divide the data set. Can I get your partition results in order to reproduce your results.

pppp-k commented 6 months ago

In addition, could you please give details of the specific evaluation indicators in the rope3d data set? Does AP in the paper refer to AP3D with easy difficulty？ Does the S of Rope is completely according to file links to https://github.com/liyingying0113/rope3d-dataset-tools/blob/main/eval_tools/run_eval_distance_v2x_ground_4categories.py? I would appreciate it if you could answer these questions！

yanglei18 commented 4 months ago

Not exactly, during the evaluation, we did not filter any GT boxes.