This jupyter notebook can be used for quick prototyping and vizualizing pointcloud frames. I use this to check output of models after training. Kitti and argoverse-kitti adapted dataset can be used. The data_seq and label_seq folders have semantic_kitti dataset pointclouds and prediction labels using Squeezeseg. The data can be vizualized using Demonstration_semantic_kitti.ipynb.
This has been also extensively used to test PointRCNN