Model-based 2D detection of unseen objects – HOT3D

This leaderbord shows the ranking for Model-based 2D detection of unseen objects on HOT3D. The reported time is the average time to generate predictions for all objects in a single image (averaged over the datasets).

Date (UTC) Submission Test image AP AP50 AP75 APS APM APL AR1 AR10 AR100 ARS ARM ARL Time (s)
2025-11-14 3PT-Detection (Nov 2025) RGB 0.484 0.586 0.565 0.036 0.418 0.546 0.616 0.670 0.675 0.122 0.586 0.772 10.987
2025-10-01 3PT-Detection (f.k.a IPT) RGB 0.481 0.583 0.560 0.040 0.411 0.542 0.610 0.656 0.657 0.099 0.561 0.764 112.256
2025-08-26 MUSE (muse_full) RGB 0.438 0.576 0.502 0.011 0.375 0.542 0.514 0.530 0.530 0.010 0.429 0.687 0.722
2024-09-20 CNOS (FastSAM) (Pyrender) RGB 0.350 0.485 0.386 0.011 0.320 0.445 0.474 0.499 0.499 0.011 0.410 0.664 0.253
2024-09-20 CNOS (SAM) (Pyrender) RGB 0.317 0.446 0.359 0.006 0.280 0.421 0.407 0.420 0.421 0.006 0.347 0.581 1.266

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

Latest submissions to this leaderboard: