1 CompactObbBlock
data_handling.CompactObbBlock(obbs, sem_id_to_name=None, probs=None)Collatable numeric OBB payload used by training and diagnostics.
obbs keeps the raw EFM ObbTW backing layout, not a simplified center size representation. The shape is (..., K, 34), where K is a padded box slot count. Valid boxes must be determined through ObbTW(obbs).get_padding_mask().
The last dimension follows the EFM layout:
0:6: object-frame 3D bounds[xmin, xmax, ymin, ymax, zmin, zmax];6:10,10:14,14:18: RGB, SLAM-left, and SLAM-right 2D boxes;18:30:PoseTWobject-to-world/object-to-snippet transform payload;30,31,32,33: semantic id, instance id, confidence, movable.
The tensor is stored as float32 even for id-like columns; use ObbTW properties such as sem_id, inst_id, prob, bb3_center_world, and bb3_diagonal when interpreting rows.
1.1 Attributes
| Name | Description |
|---|---|
| obbs | OBB tensor with shape (..., K, 34) using EFM ObbTW layout. |
| sem_id_to_name | Optional sparse semantic id to class-name map. |
| probs | Optional class probabilities aligned with obbs. |