PointRCNN
PointRCNN copied to clipboard
Why are images needed?
Hi,
I'm trying to run inference on manually collected data that does not include images and I keep on getting assertion errors where the code is checking for images. I can't find anything in defualt.yml
for disabling this behavior either.
Are the images actually needed? Based on the paper the only input should be pointcloud.
Thanks!
the images are used for filter the pointcloud
So is there a possibility to only use the pointcloud without images?
the images are used for filter the pointcloud
I want to use olny lidar of 32 lines collected by us, the network can work?
the images are used for filter the pointcloud
How to use images to filter point clouds?Is it to filter out the point cloud outside the image range?
Any findings on this thread? I too want to use only point cloud for training the network