search

GANWANSHUI / GaussianOcc

GaussianOcc: Fully Self-supervised and Efficient 3D Occupancy Estimation with Gaussian Splatting
https://ganwanshui.github.io/GaussianOcc/
Apache License 2.0
170 stars 8 forks source link

'point_cloud' in ddad dataset not found. #12

Closed synsin0 closed 2 weeks ago

synsin0 commented 2 weeks ago

Hello, I attempt to reproduce training and test on DDAD dataset and faces multiple errors. I'd like to ask about the 'point_cloud' in ddad dataset. First, the point_cloud in DDAD.tar are in .npz file but not .npy file, Then, the camera timestamp index_temporal does not match file name of point cloud. Last, the point_cloud seems not used later. Should I just neglect the point cloud file?

synsin0 commented 2 weeks ago

I eval the DDAD dataset with official checkpoint and gets the log ddad_log.txt. Is it normal for depth prediction?

GANWANSHUI commented 2 weeks ago

I eval the DDAD dataset with official checkpoint and gets the log ddad_log.txt. Is it normal for depth prediction?

Yes, it is normal though a little worse than my test, you can do some 3d occ visualization.

GANWANSHUI commented 2 weeks ago

Hello, I attempt to reproduce training and test on DDAD dataset and faces multiple errors. I'd like to ask about the 'point_cloud' in ddad dataset. First, the point_cloud in DDAD.tar are in .npz file but not .npy file, Then, the camera timestamp index_temporal does not match file name of point cloud. Last, the point_cloud seems not used later. Should I just neglect the point cloud file?

Yes, you can neglect the unused files.