Model-based 6D detection of unseen objects – HOPEv2

This leaderbord shows the ranking for Model-based 6D detection of unseen objects on HOPEv2. 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)
2025-10-01 3PT-Pose-H3 (f.k.a IPT) Single RGB-D 0.599 0.572 0.626 50.834
2025-11-14 3PT-Pose (Nov 2025) Multi RGB 0.551 0.456 0.645 15.292
2025-10-01 Co-op (MUSE, 1 Hypo) Single RGB 0.514 0.435 0.593 4.864
2025-09-18 Co-op (CNOS, 1 Hypo) Single RGB 0.477 0.405 0.549 7.484
2025-09-17 Co-op (CNOS, Coarse) (Inlier ratio) Single RGB 0.468 0.391 0.545 3.253
2024-09-23 GigaPose + GenFlow (threshold = 0.6) Single RGB 0.411 0.346 0.475 6.733
2024-11-20 OPFormer (CNOS, Megapose Refiner) Single RGB 0.392 0.337 0.447 5.774
2024-09-23 GigaPose + GenFlow Single RGB 0.358 0.302 0.413 6.184
2024-11-20 OPFormer (CNOS) Single RGB 0.351 0.272 0.431 1.190
2024-09-23 GigaPose-CVPR24 single RGB 0.167 0.046 0.288 1.138

Tip: Hover over the numbers to see more decimal places.

Latest submissions to this leaderboard: