In multi-frame models, I'm assuming you map all Lidar points to the same (egocentric) frame. Is this done during preprocessing (building of the dataset), or is it done at training time by some part of the model (e.g. right before PointPillars/VoxelNet featurization)? I could not find this part in your code. Thank you!
tianweiy
commented
2 years ago
In multi-frame models, I'm assuming you map all Lidar points to the same (egocentric) frame. Is this done during preprocessing (building of the dataset), or is it done at training time by some part of the model (e.g. right before PointPillars/VoxelNet featurization)? I could not find this part in your code. Thank you!