This table lists the benchmark results for the 3D semantic label with limited annotations scenario.




Method Infoavg ioubathtubbedbookshelfcabinetchaircountercurtaindeskdoorfloorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwallwindow
sorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
ActiveST0.703 10.977 10.776 20.657 40.707 10.874 10.541 10.744 10.605 10.610 10.968 10.442 30.126 50.705 10.785 10.742 10.791 10.586 10.940 10.839 10.645 1
Gengxin Liu, Oliver van Kaick, Hui Huang, Ruizhen Hu: Active Self-Training for Weakly Supervised 3D Scene Semantic Segmentation.
RM3Dpermissive0.662 20.812 30.762 30.742 10.635 20.828 50.474 20.736 20.588 20.546 20.947 40.450 20.174 40.536 40.752 20.668 20.735 40.583 20.902 50.797 50.573 4
Kangcheng Liu: RM3D: Robust Data-Efficient 3D Scene Parsing via Traditional and Learnt 3D Descriptors-Based Semantic Region Merging. International Journal of Computer Vision (IJCV), 2022
DE-3DLearner LA0.639 30.839 20.723 50.681 20.629 30.839 40.424 30.728 30.538 40.526 30.945 50.427 50.120 60.511 60.643 50.547 50.781 20.566 40.905 30.809 40.607 3
Ping-Chung Yu, Cheng Sun, Min Sun: Data Efficient 3D Learner via Knowledge Transferred from 2D Model. ECCV 2022
GaIA0.638 40.536 110.783 10.651 50.600 40.840 20.413 40.728 30.490 60.520 50.948 30.475 10.299 10.518 50.680 30.629 30.729 50.573 30.906 20.815 30.626 2
Min Seok Lee*, Seok Woo Yang*, and Sung Won Han: GaIA: Graphical Information gain based Attention Network for Weakly Supervised 3D Point Cloud Semantic Segmentation. WACV 2023
LE0.608 50.791 40.726 40.651 50.589 50.779 80.346 80.662 70.493 50.524 40.923 120.430 40.234 30.572 20.638 60.411 90.708 60.533 70.855 60.782 60.508 6
One-Thing-One-Click0.594 60.756 50.722 60.494 110.546 70.795 60.371 50.725 50.559 30.488 60.957 20.367 70.261 20.547 30.575 110.225 110.671 90.543 50.904 40.826 20.557 5
Zhengzhe Liu, Xiaojuan Qi, Chi-Wing Fu: One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentation. CVPR 2021
VIBUSpermissive0.586 70.736 70.623 100.664 30.559 60.840 20.358 60.666 60.447 70.429 90.944 60.421 60.000 120.411 80.629 70.614 40.745 30.541 60.848 80.758 70.493 7
Beiwen Tian,Liyi Luo,Hao Zhao,Guyue Zhou: VIBUS: Data-efficient 3D Scene Parsing with VIewpoint Bottleneck and Uncertainty-Spectrum Modeling. ISPRS Journal of Photogrammetry and Remote Sensing
PointContrast_LA_SEM0.550 80.735 80.676 70.601 80.475 80.794 70.288 100.621 90.378 110.430 80.940 70.303 90.089 90.379 90.580 100.531 60.689 80.422 100.852 70.758 70.468 8
Viewpoint_BN_LA_AIR0.548 90.747 60.574 120.631 70.456 90.762 100.355 70.639 80.412 80.404 100.940 70.335 80.107 70.277 110.645 40.495 70.666 100.517 80.818 90.740 100.431 10
Liyi Luo, Beiwen Tian, Hao Zhao, Guyue Zhou: Pointly-supervised 3D Scene Parsing with Viewpoint Bottleneck.
CSC_LA_SEM0.531 100.659 90.638 90.578 90.417 100.775 90.254 110.537 100.396 90.439 70.939 90.284 110.083 100.414 70.599 90.488 80.698 70.444 90.785 100.747 90.440 9
SQN_LA0.486 110.587 100.649 80.527 100.372 110.718 110.320 90.510 110.393 100.325 110.924 110.290 100.095 80.287 100.607 80.356 100.626 110.416 110.672 110.680 120.359 11
Scratch_LA_SEM0.382 120.389 120.606 110.401 120.303 120.705 120.169 120.460 120.292 120.282 120.939 90.207 120.004 110.147 120.201 120.184 120.592 120.389 120.409 120.714 110.250 12


This table lists the benchmark results for the 3D semantic instance with limited annotations scenario.




Method Infoavg ap 50%bathtubbedbookshelfcabinetchaircountercurtaindeskdoorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwindow
sorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
WLab3D-Net_Ins(WS3D)permissive0.548 11.000 10.690 10.476 20.406 10.756 10.031 10.733 10.215 10.351 10.415 20.319 10.541 11.000 10.477 10.576 20.557 10.941 20.377 1
Box2Mask_LA0.465 20.667 20.591 30.773 10.331 20.682 20.029 20.409 20.122 20.284 20.432 10.253 20.466 21.000 10.127 40.806 10.280 30.821 40.291 2
Julian Chibane, Francis Engelmann, Tuan Anh Tran, Gerard Pons-Moll: Box2Mask: Weakly Supervised 3D Semantic Instance Segmentation Using Bounding Boxes. ECCV 2022
CSC_LA_INS0.289 30.667 20.580 40.427 30.202 30.424 50.000 30.384 30.015 30.061 40.180 40.014 30.071 40.119 50.173 30.445 30.390 20.938 30.120 3
PointContrast_LA_INS0.259 40.333 50.286 50.334 40.142 40.485 30.000 30.343 40.010 40.127 30.219 30.005 40.324 30.267 30.226 20.402 50.103 50.994 10.069 4
Scratch_LA_INS0.200 50.667 20.673 20.145 50.100 50.430 40.000 30.314 50.004 50.025 50.099 50.000 50.000 50.143 40.076 50.424 40.198 40.297 50.006 5


This table lists the benchmark results for the 3D object detection with limited annotations scenario.




Method Infoavg ap 50%bathtubbedbookshelfcabinetchaircountercurtaindeskdoorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwindow
sorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
WS3D_LApermissive0.341 11.000 10.629 10.426 10.070 10.608 10.063 10.176 10.503 10.132 10.084 10.001 10.337 10.220 10.103 10.628 10.282 10.739 10.131 1
PointContrast_LA_DET0.135 20.667 20.296 30.145 20.002 30.330 30.000 40.000 30.050 30.049 20.023 20.000 30.002 40.006 40.034 30.472 20.052 20.285 40.011 2
CSC_LA_DET0.135 20.444 30.336 20.029 40.001 40.356 20.008 20.000 20.011 40.045 30.010 40.000 20.032 20.011 30.043 20.458 30.028 30.602 20.010 3
Scratch_LA_DET0.098 40.167 40.253 40.074 30.002 20.257 40.004 30.000 40.080 20.038 40.013 30.000 30.006 30.143 20.002 40.243 40.017 40.473 30.001 4


This table lists the benchmark results for the 3D semantic label with limited reconstructions scenario.




Method Infoavg ioubathtubbedbookshelfcabinetchaircountercurtaindeskdoorfloorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwallwindow
sorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
WeakLab3DNet (WS3D)0.682 10.863 10.765 20.782 10.648 10.803 70.438 30.793 10.607 10.589 10.944 30.455 10.223 20.536 20.768 10.726 10.758 20.623 10.906 10.821 20.596 3
MFA-Net0.678 20.779 40.782 10.774 20.637 20.827 40.491 10.736 20.597 20.561 20.947 20.438 20.206 30.610 10.758 20.667 20.773 10.594 30.880 20.824 10.673 1
DE-3DLearner LR0.608 30.853 20.689 50.593 70.483 50.830 20.466 20.652 30.528 40.482 30.954 10.288 60.250 10.448 40.595 40.532 50.748 30.503 60.822 40.806 30.647 2
Ping-Chung Yu, Cheng Sun, Min Sun: Data Efficient 3D Learner via Knowledge Transferred from 2D Model. ECCV 2022
CSC_LR_SEM0.575 40.671 80.740 30.727 30.445 60.847 10.380 70.602 50.512 50.447 50.942 40.291 50.184 40.353 80.468 80.508 60.745 40.602 20.855 30.765 50.420 8
CSG_3DSegNet0.570 50.717 60.730 40.697 40.521 30.823 50.377 80.419 80.531 30.452 40.935 80.316 30.147 50.359 70.551 70.551 40.692 70.513 50.797 60.764 60.508 4
Viewpoint_BN_LR_AIR0.566 60.780 30.659 80.677 50.484 40.799 80.419 50.636 40.480 60.432 70.940 50.238 80.124 60.396 50.609 30.432 80.735 50.527 40.787 70.752 80.423 7
PointContrast_LR_SEM0.555 70.711 70.668 60.622 60.425 70.830 20.433 40.552 60.273 80.440 60.938 60.287 70.096 70.470 30.576 50.612 30.687 80.438 80.781 80.785 40.474 5
Scratch_LR_SEM0.531 80.750 50.666 70.553 80.409 80.816 60.387 60.487 70.285 70.368 80.938 60.310 40.074 80.388 60.564 60.468 70.698 60.448 70.804 50.761 70.454 6


This table lists the benchmark results for the 3D semantic instance with limited reconstructions scenario.




Method Infoavg ap 50%bathtubbedbookshelfcabinetchaircountercurtaindeskdoorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwindow
sorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
WeakLabel3DNet(WS3D)0.649 11.000 10.800 10.721 10.603 10.807 20.044 30.735 10.377 10.466 10.550 10.605 10.550 21.000 10.506 20.776 20.618 11.000 10.526 1
InstTeacher3D0.598 21.000 10.727 50.205 60.420 20.833 10.405 10.470 30.247 20.463 20.536 20.559 20.533 31.000 10.552 10.782 10.587 21.000 10.444 2
TWIST+CSC0.481 30.667 30.760 20.468 30.313 30.802 30.008 40.529 20.098 60.364 30.411 30.348 30.500 50.571 30.504 30.646 50.530 30.944 40.201 3
CSC_LR_INS0.440 40.667 30.737 40.418 50.218 60.791 40.094 20.328 50.185 30.251 60.382 40.273 40.565 10.539 40.377 40.588 60.371 61.000 10.128 5
PointContrast_LR_INS0.432 50.667 30.757 30.560 20.278 50.740 50.003 60.435 40.123 50.309 40.347 50.109 60.522 40.429 60.223 60.739 30.434 50.944 40.149 4
Scratch_LR_INS0.413 60.667 30.720 60.442 40.288 40.735 60.005 50.326 60.138 40.302 50.329 60.204 50.445 60.498 50.229 50.657 40.452 40.889 60.115 6


This table lists the benchmark results for the 3D object detection with limited reconstructions scenario.




Method Infoavg ap 50%bathtubbedbookshelfcabinetchaircountercurtaindeskdoorotherfurniturepicturerefrigeratorshower curtainsinksofatabletoiletwindow
sorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
WLabel3DNet-LR(WS3D)permissive0.374 10.867 10.797 10.655 10.104 10.678 10.046 10.215 10.406 10.186 10.219 10.034 10.354 10.160 10.101 10.741 10.306 10.679 10.181 1
CSC_LR_DET0.191 20.667 20.468 30.226 20.036 20.420 30.025 20.010 30.081 30.066 30.045 20.000 20.162 30.010 30.017 20.657 20.109 20.420 30.013 2
PointContrast_LR_DET0.187 30.667 20.523 20.109 30.027 30.435 20.005 30.013 20.199 20.070 20.035 30.000 40.183 20.033 20.003 40.497 30.078 30.488 20.005 3
Scratch_LR_DET0.076 40.667 20.099 40.015 40.005 40.190 40.000 40.000 40.033 40.007 40.001 40.000 30.000 40.010 40.004 30.094 40.014 40.237 40.000 4