Gorilla-Lab-SCUT/frustum-convnet

How to visualize output?

Closed this issue · 5 comments

I saw the issue that was before (https://github.com/charlesq34/frustum-pointnets/tree/master/mayavi)
So I load the viz_util.py and change the data path(txt)
image
image
image

However, the index in the code is different from the one we have now so error occur.
Can you tell me how to transform the current data format? or
If you have a code to visualize the result value, could you share it?

You need to provide a point cloud data with shape(N, 3) for draw_lidar, not the detection results. If you have no idea about it, you can also reference the code https://github.com/kuixu/kitti_object_vis.

thank you for your answer

I have a question. In your model code I cant't find 2d detection (ex. rcnn f-rcnn, yolo....)
Is the detection of 2d objects a result obtained from a previous run (In kitti dataset), and is the result obtained to recognize 3d objects?