This leaderbord shows the ranking for Model-based 6D detection of unseen objects on IPD. The reported time is the average time to generate predictions for all objects in a single image (averaged over the datasets).
| Date (UTC) | Submission | Views | Test image | AP | APMSSD | APMSPD | Time (s) | |
|---|---|---|---|---|---|---|---|---|
| 2025-11-14 | 3PT-Pose (Nov 2025) | Multi | RGB | 0.984 | 0.983 | 0.986 | 24.516 | |
| 2025-09-29 | FRTPose-WAPR.v2（Multi View） | Multi | RGB-D | 0.926 | 0.891 | 0.961 | 93.013 | |
| 2025-10-01 | 3PT-Pose-Industrial (f.k.a. IPT) | Multi | RGB | 0.923 | 0.914 | 0.933 | 160.481 | |
| 2025-06-01 | FRTPose-WAPR（Multi Cam） | Multi | RGB-D | 0.918 | 0.881 | 0.954 | 198.809 | |
| 2025-09-30 | FRTPose-WAPR.v2（Default Detection） | Single | RGB-D | 0.608 | 0.596 | 0.621 | 27.400 | |
| 2025-06-01 | FRTPose-WAPR（Default Detection） | Single | RGB-D | 0.583 | 0.567 | 0.599 | 75.059 | |
| 2025-10-01 | FRTPose-WAPR.v2（SAM6D / MUSE） (sam6d 0.4) | Single | RGB-D | 0.537 | 0.526 | 0.547 | 0.931 | |
| 2025-07-17 | MatchU | Single | RGB-D | 0.520 | 0.475 | 0.566 | -1.000 | |
| 2025-10-01 | mask matcher (Flanders Make) | Multi | RGB-D | 0.496 | 0.364 | 0.628 | 41.170 | |
| 2025-05-30 | FRTPose-WAPR（SAM6D / MUSE） | Single | RGB-D | 0.481 | 0.469 | 0.493 | 0.764 | |
| 2025-10-01 | mask matcher (Flanders Make) | Multi | RGB-D | 0.474 | 0.353 | 0.595 | 41.319 | |
| 2025-10-01 | mask matcher | Multi | RGB-D | 0.468 | 0.341 | 0.594 | 20.717 | |
| 2025-05-31 | FreeZeV2.2 | Single | RGB-D | 0.439 | 0.419 | 0.459 | 7.201 | |
| 2025-06-11 | SAM6D(SAM) | Single | RGB-D | 0.415 | 0.391 | 0.440 | 2.414 | |
| 2025-06-12 | FoundationPose(CNOS-SAM) | Single | RGB-D | 0.408 | 0.392 | 0.424 | 19.021 |
Tip: Hover over the numbers to see more decimal places.