# OmniPart + TRELLIS2 Assembly Diagnostics

- Run: `/data/250010098/physxanything_stage1_experiments/runs/omnipart_trellis2_perpart_formal_softfit_selected_20260417/0023687e90394c3e97ab19b0160cafb3`
- Success parts: `2 / 2`
- OmniPart boxes: `8`

## Part Stats

| idx | part_id | crop | target volume | mask px | faces | mesh/target IoU | fill ratio |
| --- | --- | --- | ---: | ---: | ---: | ---: | ---: |
| 0 | 4 | dim_context | 0.005147 | 5374 | 7088140 | 0.8342 | 0.8342 |
| 1 | 1 | dim_context | 0.005095 | 4103 | 631754 | 0.7378 | 0.8269 |

## Pairwise Target BBox Relations

| part_i | part_j | bbox IoU | intersection volume | gap |
| --- | --- | ---: | ---: | ---: |
| 4 | 1 | 0.0000 | 0.000000 | 0.079365 |

## Notes

- These diagnostics are AABB-level sanity checks, not surface-intersection tests.
- High target bbox IoU means assembly conditions overlap before any mesh generation.
- Low mesh-to-target bbox IoU or very anisotropic extent ratios can indicate poor bbox fit or part hallucination.