sangjin

Hello. Thank you for sharing your nice project here. I wonder what the `Distribution` class of `ae.py` in `models` means. What does this class imply?

No module named 'cumm' when building spconv from source on AGX Orin

2

Hi, I run this commands ```shell export CUMM_CUDA_ARCH_LIST="8.7" git clone https://github.com/FindDefinition/cumm cd cumm pip install -e . git clone https://github.com/traveller59/spconv cd spconv ``` And I've removed the `cumm` in `requires`...

Hi, thank you for the awesome work!! Your repo is greatly helpful to me!! I have two questions. 1) The tracking time exponentially increases with respect to the number of...

Image BEV feature was getting worse during training

1

Hi I'm trying to train the BEVFusion using a single image and its corresponding point cloud for my custom dataset. But, the performance gap was not large between LiDAR and...

When using single image (e.g. KITTI)

32

Hi, thank you for your great work! I'm trying to learn BEVFusion on a KITT dataset. Since using a single image, I have used 5 images as the zero array....

incorrect inference on KITTI dataset

1

Hi, I trained BEVFusion on KITTI dataset and I got a reasonable performance with the naive torch model. I have successfully built FP16 and INT8 models. But both models produce...

Does the point cloud range need to be symmetrical from zero?

1

Hi, When I use `point_cloud_range=[-54, -100, -5 54, 100, 3]`, the prediction results are quite reasonable. - `point_cloud_range=[-54, -100, -5 54, 100, 3]` - `voxel_size=[0.075, 0.1389, 0.2]` However, in the...