About the accuracy of PDV on Kitti test sets

[S-Eternal-S]

Hello, authors. The ground truth of the kitti test set is not available, how do you calculate the accuracy of the network on the test set?

[jskhu]

You can evaluate it on the KITTI online benchmark: https://www.cvlibs.net/datasets/kitti/eval_object.php?obj_benchmark=3d. You can generate the results for the benchmarks by specifying the --save_to_file command in tools/test.py: https://github.com/TRAILab/PDV/blob/024a3695cae782adf7a5fa12c34b0ad461e0844c/tools/test.py#L42

[S-Eternal-S]

Thank you very much!