MBPTrack3D

MBPTrack3D copied to clipboard

No module named 'pointnet2_ops._ext'

2

   Hello author, I tried using the same environment as you but encountered the problem in the picture. How did you solve it

bobo-liang

No module named 'pointnet2_ops._ext'

1

Hello author, when I ran the MBPTrack3D project in the conda environment of the CXTrack3D project, the following problem occurred, but I think the environments of the two projects should...

newbie521

Reproduction of the result.

6

Dear authors, I am not able to reproduce the results reported in the paper using the given config at config/mbptrack_kitti_ped_cfg.yaml. I wonder if the same config used for generating the...

IzhiSu

Little or no change in the predicted value of the car

1

When I used mbptrack_nuscenes_kitti_car.ckpt to predict the nuscenes-mini data set, the predicted value changed very little or I got a result with no change at all. for example: 675.94 1629.18...

learnuser1

Coordinate_mode "velodyne" and "camera".

1

Hello, thanks for your wonderful work. I have a question: When I switch the coordinate_mode to "velodyne", the testing results **Success** always show a value of 2.500. How can I...

hushbuddy

In the nuScenes dataset, which includes camera, LiDAR, and radar data, does the model only require the LiDAR data for training, specifically the LIDAR_TOP files in the sweeps and samples...

ljg-101

Question About Waymo Dataset Download

8

I visited the [Waymo Open Dataset download page](https://waymo.com/open/download/) to download the dataset, but there are too many versions, and I'm not sure which one to download. From what I can...

ujaejoon

Reproduction of results

3

Hi, Thanks for your efforts in this publication. It seems that the results are heavily affected by the version of `pytorch `and `cuda`. For instance, when using `torch 1.12.1` and...

StrivedTye

**环境是：** - open3d 0.15.2 - python 3.8.18 - pytorch 1.10.0(cuda11.3,cudnn8.2) - pytorch-lightning 1.5.10 - pytorch3d 0.6.2 - shapely 1.7.1 - torchvision 0.11.0+cu113 **仅仅因为显存原因调整了batch_size为8，训练时异常：** > model.backward(closure_loss, optimizer, *args, **kwargs) > File...

Hichengdong

Question Regarding Bounding Box Orientation from KITTI Dataset Labels in Camera Coordinates

2

 This is code from https://github.com/slothfulxtx/MBPTrack3D/blob/main/datasets/kitti_mem.py As I understand, in the KITTI dataset, the bounding box orientation is already annotated in the camera coordinate system (Reference: [KITTI Coordinate Transformations](https://towardsdatascience.com/kitti-coordinate-transformations-125094cd42fb)). Given...

ujaejoon