search

duanyiqun / DiffusionDepth

PyTorch Implementation of introducing diffusion approach to 3D depth perception
https://arxiv.org/abs/2303.05021
Apache License 2.0
293 stars 16 forks source link

About kitti eigen split #42

Closed wangjiyuan9 closed 10 months ago

wangjiyuan9 commented 10 months ago

Hi there, thank you for your amazing work! I noticed that in the paper, you mentioned the eigensplit you had used. However, I never found any eigen train split file like other projects. And you said to use prepare_KITTI_DP to get the kitti data but never said how to get gt (for depth annotated data, we had gt, but eigen split != annotated one)

wangjiyuan9 commented 10 months ago

To be specific, there had said the difference between annotated depth and the projective point depth.

wangjiyuan9 commented 10 months ago

And actually, I found no where to get the gt depth of the eigen train split, which is necessary for training :(

duanyiqun commented 10 months ago

The prepare_KITTI_DP is for official split. Eigen train split is a widely used train split on Kitti dataset by many papers. And actually, you may have to check previous papers for Eigen's paper and split by yourself. Regards

wangjiyuan9 commented 10 months ago

Thank you for your reply! Yes, I know that the Eigen train split is a traditional split, but how to get the gt_depth which is training necessary? Did you do it as monodepth2??