fregu856 / ebms_3dod Goto Github PK

The 3dod/viz_video.py, which is the code for creating the Youtube video with qualitative results can not be run successfully, please update it with corrections, thanks

@fregu856 any tentative dates for code release??

Hi, when will the code be released?

Thanks for sharing. @fregu856
You apply EBM-based regression to 3DOD by extending the SA-SSD with a conditional EBM p(y|x; θ). Why don't you just use a DNN f(θ) to predict the un-normalized density directly from the input-target pair (x, y) like in 2D Object Detection? If so you don't need to extract BEV features in SA-SSD and design a differentiable pooling operator.

The ubuntu 18.04 can not support spconv 1.0 anymore, please update it, thanks!

Hi,

I am working on your code to further develop the idea, I would like to know, for only testing the pre-trained model which dataset I must download? Only the Lidar dataset in the Kitti-3d object? I am only interested in point cloud, and not sensor fusion with camera.
I attached a photo to this issue, and I just want to confirm which dataset, or what set of datasets do I need to test the pre-trained model on.

Thank you for your help.