Whats the difference between feeding 2-d point clouds and 2-d images to a CNN
Can a projection of 3 d point cloud to a 2d space be fed to a CNN in the same way that a normal 2-d image is fed in? In terms of model and architecture what would need to be changed?
Asked by DL_geek on 2020-05-16 09:21:49 UTC
Comments
This does not seem to be a ROS question.
Asked by gvdhoorn on 2020-05-16 11:09:55 UTC