Benchmark Results - ScanNet Benchmark

This table lists the benchmark results for the ScanNet200 3D semantic label scenario.

Method	avg iou	head iou	common iou	tail iou	backpack	bag	ball	bar	basket	bathroom cabinet	bathroom stall	bathroom stall door	bathroom vanity	bathtub	bed	bench	bicycle	bin	blackboard	blanket	blinds	board	book	bookshelf	bottle	bowl	box	broom	bucket	bulletin board	cabinet	calendar	candle	cart	case of water bottles	ceiling	ceiling light	chair	clock	closet	closet rod	clothes	clothes dryer	coat rack	coffee kettle	coffee maker	coffee table	column	computer tower	copier	couch	counter	cup	curtain	cushion	decoration	desk	dining table	dish rack	dishwasher	door	doorframe	dresser	dumbbell	end table	fan	file cabinet	fire alarm	fire extinguisher	floor	folded chair	hat	headphones	ironing board	jacket	keyboard	keyboard piano	kitchen cabinet	kitchen counter	ladder	lamp	laptop	laundry basket	laundry detergent	laundry hamper	ledge	light	light switch	machine	mailbox	mat	microwave	mini fridge	mirror	monitor	mouse	music stand	nightstand	object	office chair	oven	paper	paper bag	paper cutter	paper towel dispenser	paper towel roll	person	piano	picture	pillow	pipe	plant	plate	plunger	poster	power outlet	power strip	printer	projector	projector screen	rack	radiator	range hood	recycling bin	refrigerator	scale	seat	shelf	shoe	shower	shower curtain	shower curtain rod	shower floor	shower head	shower wall	sign	sink	soap dispenser	sofa chair	speaker	stair rail	stairs	stand	stool	storage bin	storage container	storage organizer	stove	structure	stuffed animal	suitcase	table	telephone	tissue box	toaster	toilet	toilet paper	toilet paper dispenser	towel	trash bin	trash can	tray	tube	tv	tv stand	vacuum cleaner	vent	wall	wardrobe	washing machine	water bottle	water cooler	water pitcher	whiteboard	window	windowsill

CeCo	0.340 1	0.551 1	0.247 1	0.181 1	0.475 2	0.057 4	0.142 3	0.000 1	0.000 1	0.000 1	0.387 2	0.463 1	0.499 2	0.924 1	0.774 1	0.213 1	0.257 1	0.000 3	0.546 4	0.100 2	0.006 2	0.615 1	0.177 4	0.534 1	0.246 1	0.000 2	0.400 1	0.000 1	0.338 1	0.006 3	0.484 1	0.609 1	0.000 1	0.083 1	0.000 2	0.873 1	0.089 2	0.661 2	0.000 3	0.048 4	0.560 1	0.408 1	0.892 1	0.000 1	0.000 1	0.586 1	0.616 2	0.000 4	0.692 2	0.900 1	0.721 1	0.162 1	0.228 1	0.860 1	0.000 1	0.000 2	0.575 1	0.083 2	0.550 1	0.347 1	0.624 1	0.410 1	0.360 1	0.740 1	0.109 2	0.321 2	0.660 1	0.000 2	0.121 2	0.939 1	0.143 2	0.000 1	0.400 1	0.003 2	0.190 1	0.564 1	0.652 1	0.615 1	0.421 1	0.304 3	0.579 1	0.547 1	0.000 1	0.000 1	0.296 1	0.000 4	0.030 4	0.096 1	0.000 2	0.916 1	0.037 1	0.551 1	0.171 2	0.376 1	0.865 1	0.286 1	0.000 1	0.633 1	0.102 4	0.027 4	0.011 2	0.000 1	0.000 1	0.474 2	0.742 1	0.133 2	0.311 1	0.824 1	0.242 1	0.503 1	0.068 3	0.828 1	0.000 2	0.429 1	0.000 1	0.063 1	0.000 1	0.781 1	0.000 1	0.000 2	0.000 1	0.665 1	0.633 1	0.450 1	0.818 1	0.000 1	0.000 1	0.429 1	0.532 1	0.226 1	0.825 1	0.510 3	0.377 1	0.709 1	0.079 2	0.000 1	0.753 1	0.683 1	0.102 4	0.063 2	0.401 4	0.620 3	0.000 1	0.619 1	0.000 4	0.000 3	0.000 1	0.595 2	0.000 2	0.000 1	0.345 2	0.564 1	0.411 1	0.603 1	0.384 1	0.945 1	0.266 1	0.643 1	0.367 1	0.304 1	0.663 1	0.000 1	0.010 1	0.726 2	0.767 1	0.898 1	0.000 1	0.784 1	0.435 1	0.861 1	0.000 1	0.447 1	0.000 4	0.257 1	0.656 1	0.377 3
: Understanding Imbalanced Semantic Segmentation Through Neural Collapse.
LGround	0.272 2	0.485 2	0.184 2	0.106 2	0.476 1	0.077 2	0.218 1	0.000 1	0.000 1	0.000 1	0.547 1	0.295 2	0.540 1	0.746 3	0.745 2	0.058 3	0.112 4	0.005 1	0.658 2	0.077 4	0.000 3	0.322 2	0.178 3	0.512 3	0.190 2	0.199 1	0.277 2	0.000 1	0.000 2	0.173 1	0.399 2	0.000 2	0.000 1	0.039 3	0.000 2	0.858 2	0.085 3	0.676 1	0.002 1	0.103 1	0.498 2	0.323 2	0.703 2	0.000 1	0.000 1	0.296 2	0.549 3	0.216 1	0.702 1	0.768 2	0.718 2	0.028 3	0.092 3	0.786 3	0.000 1	0.000 2	0.453 3	0.022 3	0.251 4	0.252 2	0.572 2	0.348 2	0.321 2	0.514 2	0.063 3	0.279 3	0.552 2	0.000 2	0.019 3	0.932 2	0.132 3	0.000 1	0.000 2	0.000 4	0.156 4	0.457 2	0.623 2	0.518 2	0.265 3	0.358 2	0.381 2	0.395 2	0.000 1	0.000 1	0.127 4	0.012 3	0.051 1	0.000 2	0.000 2	0.886 3	0.014 2	0.437 4	0.179 1	0.244 2	0.826 2	0.000 2	0.000 1	0.599 2	0.136 1	0.085 2	0.000 3	0.000 1	0.000 1	0.565 1	0.612 2	0.143 1	0.207 2	0.566 2	0.232 3	0.446 2	0.127 1	0.708 3	0.000 2	0.384 2	0.000 1	0.000 2	0.000 1	0.402 2	0.000 1	0.059 1	0.000 1	0.525 4	0.566 2	0.229 3	0.659 3	0.000 1	0.000 1	0.265 2	0.446 2	0.147 3	0.720 4	0.597 2	0.066 3	0.000 2	0.187 1	0.000 1	0.726 2	0.467 4	0.134 3	0.000 3	0.413 3	0.629 2	0.000 1	0.363 3	0.055 3	0.022 2	0.000 1	0.626 1	0.000 2	0.000 1	0.323 3	0.479 4	0.154 3	0.117 2	0.028 3	0.901 2	0.243 2	0.415 4	0.295 4	0.143 4	0.610 3	0.000 1	0.000 3	0.777 1	0.397 4	0.324 3	0.000 1	0.778 2	0.179 2	0.702 3	0.000 1	0.274 4	0.404 1	0.233 2	0.622 2	0.398 2
David Rozenberszki, Or Litany, Angela Dai: Language-Grounded Indoor 3D Semantic Segmentation in the Wild. arXiv
Minkowski 34D	0.253 3	0.463 3	0.154 4	0.102 3	0.381 4	0.084 1	0.134 4	0.000 1	0.000 1	0.000 1	0.386 3	0.141 4	0.279 4	0.737 4	0.703 3	0.014 4	0.164 3	0.000 3	0.663 1	0.092 3	0.000 3	0.224 3	0.291 1	0.531 2	0.056 4	0.000 2	0.242 3	0.000 1	0.000 2	0.013 2	0.331 3	0.000 2	0.000 1	0.035 4	0.001 1	0.858 2	0.059 4	0.650 4	0.000 3	0.056 3	0.353 3	0.299 3	0.670 3	0.000 1	0.000 1	0.284 3	0.484 4	0.071 3	0.594 3	0.720 3	0.710 3	0.027 4	0.068 4	0.813 2	0.000 1	0.005 1	0.492 2	0.164 1	0.274 3	0.111 4	0.571 3	0.307 4	0.293 3	0.307 4	0.150 1	0.163 4	0.531 3	0.002 1	0.545 1	0.932 2	0.093 4	0.000 1	0.000 2	0.002 3	0.159 3	0.368 4	0.581 4	0.440 4	0.228 4	0.406 1	0.282 4	0.294 3	0.000 1	0.000 1	0.189 3	0.060 1	0.036 3	0.000 2	0.000 2	0.897 2	0.000 4	0.525 3	0.025 4	0.205 4	0.771 4	0.000 2	0.000 1	0.593 3	0.108 3	0.044 3	0.000 3	0.000 1	0.000 1	0.282 4	0.589 3	0.094 3	0.169 3	0.466 4	0.227 4	0.419 4	0.125 2	0.757 2	0.002 1	0.334 3	0.000 1	0.000 2	0.000 1	0.357 3	0.000 1	0.000 2	0.000 1	0.582 2	0.513 4	0.337 2	0.612 4	0.000 1	0.000 1	0.250 3	0.352 4	0.136 4	0.724 3	0.655 1	0.280 2	0.000 2	0.046 4	0.000 1	0.606 4	0.559 2	0.159 1	0.102 1	0.445 1	0.655 1	0.000 1	0.310 4	0.117 1	0.000 3	0.000 1	0.581 4	0.026 1	0.000 1	0.265 4	0.483 3	0.084 4	0.097 4	0.044 2	0.865 4	0.142 4	0.588 2	0.351 2	0.272 2	0.596 4	0.000 1	0.003 2	0.622 3	0.720 2	0.096 4	0.000 1	0.771 3	0.016 3	0.772 2	0.000 1	0.302 3	0.194 2	0.214 3	0.621 3	0.197 4
C. Choy, J. Gwak, S. Savarese: 4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks. CVPR 2019
CSC-Pretrain	0.249 4	0.455 4	0.171 3	0.079 4	0.418 3	0.059 3	0.186 2	0.000 1	0.000 1	0.000 1	0.335 4	0.250 3	0.316 3	0.766 2	0.697 4	0.142 2	0.170 2	0.003 2	0.553 3	0.112 1	0.097 1	0.201 4	0.186 2	0.476 4	0.081 3	0.000 2	0.216 4	0.000 1	0.000 2	0.001 4	0.314 4	0.000 2	0.000 1	0.055 2	0.000 2	0.832 4	0.094 1	0.659 3	0.002 1	0.076 2	0.310 4	0.293 4	0.664 4	0.000 1	0.000 1	0.175 4	0.634 1	0.130 2	0.552 4	0.686 4	0.700 4	0.076 2	0.110 2	0.770 4	0.000 1	0.000 2	0.430 4	0.000 4	0.319 2	0.166 3	0.542 4	0.327 3	0.205 4	0.332 3	0.052 4	0.375 1	0.444 4	0.000 2	0.012 4	0.930 4	0.203 1	0.000 1	0.000 2	0.046 1	0.175 2	0.413 3	0.592 3	0.471 3	0.299 2	0.152 4	0.340 3	0.247 4	0.000 1	0.000 1	0.225 2	0.058 2	0.037 2	0.000 2	0.207 1	0.862 4	0.014 2	0.548 2	0.033 3	0.233 3	0.816 3	0.000 2	0.000 1	0.542 4	0.123 2	0.121 1	0.019 1	0.000 1	0.000 1	0.463 3	0.454 4	0.045 4	0.128 4	0.557 3	0.235 2	0.441 3	0.063 4	0.484 4	0.000 2	0.308 4	0.000 1	0.000 2	0.000 1	0.318 4	0.000 1	0.000 2	0.000 1	0.545 3	0.543 3	0.164 4	0.734 2	0.000 1	0.000 1	0.215 4	0.371 3	0.198 2	0.743 2	0.205 4	0.062 4	0.000 2	0.079 2	0.000 1	0.683 3	0.547 3	0.142 2	0.000 3	0.441 2	0.579 4	0.000 1	0.464 2	0.098 2	0.041 1	0.000 1	0.590 3	0.000 2	0.000 1	0.373 1	0.494 2	0.174 2	0.105 3	0.001 4	0.895 3	0.222 3	0.537 3	0.307 3	0.180 3	0.625 2	0.000 1	0.000 3	0.591 4	0.609 3	0.398 2	0.000 1	0.766 4	0.014 4	0.638 4	0.000 1	0.377 2	0.004 3	0.206 4	0.609 4	0.465 1
Ji Hou, Benjamin Graham, Matthias Nießner, Saining Xie: Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene Contexts. CVPR 2021

This table lists the benchmark results for the ScanNet200 3D semantic instance scenario.

Method	avg	alarm clock	armchair	backpack	bar	basket	bathroom cabinet	bathroom counter	bathroom stall	bathroom stall door	bathtub	bed	bench	bicycle	bin	blackboard	blanket	blinds	board	book	bookshelf	bottle	bowl	box	broom	bucket	bulletin board	cabinet	calendar	candle	cart	case of water bottles	cd case	ceiling	ceiling light	clock	closet	closet door	closet rod	closet wall	clothes dryer	coffee kettle	coffee maker	coffee table	column	computer tower	container	copier	couch	crate	cup	curtain	decoration	desk	dining table	dish rack	dishwasher	divider	door	doorframe	dumbbell	dustpan	end table	fan	fire alarm	fire extinguisher	fireplace	folded chair	furniture	guitar case	jacket	keyboard	keyboard piano	kitchen cabinet	kitchen counter	ladder	lamp	laptop	laundry basket	laundry detergent	laundry hamper	ledge	light	light switch	luggage	machine	mailbox	mattress	microwave	mini fridge	monitor	mouse	music stand	nightstand	object	office chair	ottoman	oven	paper	paper cutter	paper towel dispenser	paper towel roll	person	piano	picture	pillar	pillow	pipe	plate	plunger	poster	potted plant	power outlet	power strip	projector	projector screen	purse	rack	radiator	range hood	recycling bin	scale	seat	shelf	shoe	shower	shower curtain	shower curtain rod	shower door	shower floor	shower head	sign	sink	soap dish	soap dispenser	sofa chair	stair rail	stairs	stand	stool	storage bin	storage container	storage organizer	stove	structure	stuffed animal	suitcase	table	telephone	tissue box	toaster	toaster oven	toilet	toilet paper	toilet paper holder	toilet seat cover dispenser	towel	tray	tube	tv	tv stand	vacuum cleaner	vent	wardrobe	washing machine	water bottle	water cooler	water pitcher	whiteboard	window	windowsill

Mask3D Scannet200	0.388 1	0.542 1	0.357 1	0.237 1	0.610 1	0.091 1	0.125 4	0.000 1	0.000 1	0.000 1	0.065 2	0.668 1	0.451 1	1.000 1	0.955 1	0.640 1	0.500 1	0.039 1	0.125 2	0.063 1	0.409 1	0.311 1	0.291 1	0.609 2	0.266 1	0.000 1	0.163 1	0.000 1	0.008 1	0.044 1	0.496 1	1.000 1	0.000 1	0.018 1	0.000 1	0.756 1	0.573 1	0.808 1	0.000 1	0.010 1	0.042 2	0.130 2	0.552 1	0.042 1	0.000 1	1.000 1	0.725 3	0.750 1	0.883 1	1.000 1	0.832 3	0.024 1	0.107 1	0.614 2	0.226 1	0.250 1	0.628 1	0.792 1	0.677 2	0.400 1	0.741 1	0.278 1	0.511 1	0.077 4	0.111 1	0.313 1	0.715 1	0.302 1	0.017 2	0.200 1	0.000 1	0.188 1	0.000 1	0.178 1	0.736 1	1.000 1	0.615 1	0.514 1	0.409 1	0.380 4	0.600 1	0.000 1	0.000 1	0.400 1	0.013 1	0.254 1	0.381 1	0.000 1	0.123 3	0.400 1	0.839 1	0.258 1	0.463 1	0.926 1	0.265 1	0.000 1	0.857 1	0.099 1	0.021 1	0.500 1	0.027 1	0.028 1	1.000 1	0.502 4	0.016 1	0.076 3	0.500 1	0.612 1	0.578 1	0.005 1	0.597 1	0.194 1	0.497 1	0.000 1	0.500 1	0.000 1	0.323 3	0.000 1	1.000 1	0.000 1	0.748 1	0.708 2	0.050 3	0.890 1	1.000 1	0.008 1	0.151 2	0.301 1	1.000 1	1.000 1	0.792 2	0.945 1	1.000 1	0.511 1	0.004 1	0.753 1	0.776 1	0.287 1	0.020 1	0.003 3	0.974 2	0.033 1	0.412 4	0.000 1	0.000 1	0.000 1	0.667 1	0.000 1	0.000 1	0.491 1	0.676 1	0.352 1	0.335 1	0.060 1	0.822 4	0.527 1	1.000 1	0.517 1	0.606 1	0.853 1	0.000 1	0.004 1	0.806 1	1.000 1	0.727 1	0.000 1	0.042 1	0.739 1	0.000 1	0.399 2	0.391 1	0.504 1	0.591 1	0.571 1
Jonas Schult, Francis Engelmann, Alexander Hermans, Or Litany, Siyu Tang, Bastian Leibe: Mask3D for 3D Semantic Instance Segmentation. ICRA 2023
LGround Inst.	0.246 2	0.413 2	0.170 2	0.130 2	0.455 4	0.003 4	0.500 1	0.000 1	0.000 1	0.000 1	0.017 3	0.333 3	0.111 4	1.000 1	0.681 3	0.400 2	0.000 2	0.000 2	1.000 1	0.003 4	0.000 2	0.167 2	0.190 2	0.637 1	0.067 2	0.000 1	0.081 2	0.000 1	0.000 2	0.000 2	0.264 3	0.000 2	0.000 1	0.000 2	0.000 1	0.387 3	0.031 4	0.754 2	0.000 1	0.000 2	0.151 1	0.135 1	0.056 3	0.000 2	0.000 1	0.582 3	0.589 4	0.500 2	0.815 2	1.000 1	0.903 1	0.000 2	0.097 2	0.588 3	0.000 2	0.000 2	0.234 2	0.000 2	0.500 3	0.400 1	0.682 3	0.156 2	0.159 3	0.750 1	0.046 2	0.125 3	0.660 2	0.000 2	0.200 1	0.000 4	0.000 1	0.000 2	0.000 1	0.164 2	0.402 2	0.500 2	0.373 2	0.025 2	0.143 4	0.426 2	0.317 2	0.000 1	0.000 1	0.000 2	0.000 2	0.063 2	0.000 2	0.000 1	0.000 4	0.000 3	0.575 3	0.250 2	0.241 2	0.772 2	0.000 2	0.000 1	0.653 3	0.034 2	0.000 2	0.000 2	0.000 2	0.000 2	1.000 1	0.561 3	0.000 2	0.100 2	0.500 1	0.541 4	0.452 2	0.000 2	0.581 2	0.000 2	0.364 2	0.000 1	0.000 2	0.000 1	0.571 1	0.000 1	0.000 2	0.000 1	0.568 4	0.511 3	0.167 2	0.857 2	0.000 2	0.000 2	0.164 1	0.112 2	0.000 4	0.530 4	1.000 1	0.286 2	0.000 2	0.125 2	0.000 2	0.464 4	0.706 2	0.208 3	0.000 2	0.125 1	0.744 3	0.000 2	0.500 1	0.000 1	0.000 1	0.000 1	0.511 2	0.000 1	0.000 1	0.344 2	0.541 2	0.068 2	0.333 2	0.000 2	1.000 1	0.196 3	0.533 3	0.318 2	0.000 3	0.748 3	0.000 1	0.000 2	0.690 2	1.000 1	0.400 3	0.000 1	0.000 2	0.667 2	0.000 1	0.333 3	0.333 2	0.270 3	0.399 2	0.083 3
David Rozenberszki, Or Litany, Angela Dai: Language-Grounded Indoor 3D Semantic Segmentation in the Wild.
CSC-Pretrain Inst.	0.209 3	0.361 4	0.157 3	0.085 3	0.506 2	0.007 2	0.500 1	0.000 1	0.000 1	0.000 1	0.000 4	0.093 4	0.221 3	0.667 3	0.524 4	0.400 2	0.000 2	0.000 2	0.000 3	0.004 3	0.000 2	0.000 4	0.109 4	0.589 3	0.000 3	0.000 1	0.059 4	0.000 1	0.000 2	0.000 2	0.322 2	0.000 2	0.000 1	0.000 2	0.000 1	0.405 2	0.055 3	0.700 4	0.000 1	0.000 2	0.028 3	0.091 4	0.083 2	0.000 2	0.000 1	0.667 2	0.768 1	0.000 3	0.807 3	1.000 1	0.776 4	0.000 2	0.000 4	0.340 4	0.000 2	0.000 2	0.103 4	0.000 2	0.750 1	0.200 3	0.634 4	0.053 4	0.246 2	0.677 2	0.006 4	0.198 2	0.432 3	0.000 2	0.000 3	0.050 3	0.000 1	0.000 2	0.000 1	0.111 4	0.356 3	0.500 2	0.188 4	0.000 3	0.220 3	0.448 1	0.050 4	0.000 1	0.000 1	0.000 2	0.000 2	0.032 4	0.000 2	0.000 1	0.396 1	0.000 3	0.573 4	0.000 4	0.228 3	0.747 3	0.000 2	0.000 1	0.573 4	0.021 4	0.000 2	0.000 2	0.000 2	0.000 2	0.500 3	0.573 2	0.000 2	0.000 4	0.125 4	0.592 3	0.364 4	0.000 2	0.450 4	0.000 2	0.364 2	0.000 1	0.000 2	0.000 1	0.340 2	0.000 1	0.000 2	0.000 1	0.610 3	0.833 1	0.221 1	0.702 4	0.000 2	0.000 2	0.135 4	0.094 3	0.125 2	0.571 3	0.500 3	0.143 4	0.000 2	0.125 2	0.000 2	0.618 2	0.667 3	0.115 4	0.000 2	0.125 1	1.000 1	0.000 2	0.500 1	0.000 1	0.000 1	0.000 1	0.502 3	0.000 1	0.000 1	0.312 3	0.248 4	0.050 3	0.000 4	0.000 2	0.997 2	0.420 2	0.500 4	0.149 4	0.451 2	0.748 2	0.000 1	0.000 2	0.636 3	0.667 4	0.600 2	0.000 1	0.000 2	0.278 4	0.000 1	0.333 3	0.000 4	0.294 2	0.381 4	0.110 2
Ji Hou, Benjamin Graham, Matthias Nießner, Saining Xie: Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene Contexts. CVPR 2021
Minkowski 34D Inst.	0.203 4	0.369 3	0.134 4	0.078 4	0.479 3	0.003 3	0.500 1	0.000 1	0.000 1	0.000 1	0.100 1	0.371 2	0.300 2	0.667 3	0.746 2	0.400 2	0.000 2	0.000 2	0.000 3	0.031 2	0.000 2	0.074 3	0.165 3	0.413 4	0.000 3	0.000 1	0.070 3	0.000 1	0.000 2	0.000 2	0.221 4	0.000 2	0.000 1	0.000 2	0.000 1	0.372 4	0.070 2	0.706 3	0.000 1	0.000 2	0.000 4	0.123 3	0.033 4	0.000 2	0.000 1	0.422 4	0.732 2	0.000 3	0.778 4	1.000 1	0.845 2	0.000 2	0.090 3	0.636 1	0.000 2	0.000 2	0.158 3	0.000 2	0.250 4	0.050 4	0.693 2	0.123 3	0.051 4	0.385 3	0.009 3	0.118 4	0.406 4	0.000 2	0.000 3	0.200 1	0.000 1	0.000 2	0.000 1	0.133 3	0.307 4	0.500 2	0.251 3	0.000 3	0.281 2	0.402 3	0.317 2	0.000 1	0.000 1	0.000 2	0.000 2	0.060 3	0.000 2	0.000 1	0.396 1	0.200 2	0.669 2	0.021 3	0.218 4	0.720 4	0.000 2	0.000 1	0.696 2	0.025 3	0.000 2	0.000 2	0.000 2	0.000 2	0.125 4	0.596 1	0.000 2	0.191 1	0.500 1	0.595 2	0.369 3	0.000 2	0.500 3	0.000 2	0.143 4	0.000 1	0.000 2	0.000 1	0.226 4	0.000 1	0.000 2	0.000 1	0.701 2	0.511 3	0.000 4	0.851 3	0.000 2	0.000 2	0.150 3	0.052 4	0.100 3	0.981 2	0.500 3	0.286 2	0.000 2	0.000 4	0.000 2	0.545 3	0.522 4	0.250 2	0.000 2	0.000 4	0.522 4	0.000 2	0.500 1	0.000 1	0.000 1	0.000 1	0.282 4	0.000 1	0.000 1	0.178 4	0.382 3	0.018 4	0.056 3	0.000 2	0.997 2	0.107 4	0.677 2	0.313 3	0.000 3	0.726 4	0.000 1	0.000 2	0.583 4	0.903 3	0.200 4	0.000 1	0.000 2	0.333 3	0.000 1	0.442 1	0.083 3	0.109 4	0.387 3	0.000 4
C. Choy, J. Gwak, S. Savarese: 4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks. CVPR 2019

ScanNet Benchmark

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

Method	avg iou	bathtub	bed	bookshelf	cabinet	chair	counter	curtain	desk	door	floor	otherfurniture	picture	refrigerator	shower curtain	sink	sofa	table	toilet	wall	window

Mix3D	0.781 1	0.964 1	0.855 1	0.843 10	0.781 1	0.858 7	0.575 3	0.831 17	0.685 6	0.714 1	0.979 1	0.594 3	0.310 15	0.801 1	0.892 8	0.841 2	0.819 3	0.723 3	0.940 7	0.887 1	0.725 12
Alexey Nekrasov, Jonas Schult, Or Litany, Bastian Leibe, Francis Engelmann: Mix3D: Out-of-Context Data Augmentation for 3D Scenes. 3DV 2021 (Oral)
O-CNN	0.762 4	0.924 2	0.823 4	0.844 9	0.770 2	0.852 10	0.577 2	0.847 14	0.711 1	0.640 14	0.958 9	0.592 4	0.217 55	0.762 10	0.888 9	0.758 8	0.813 6	0.726 1	0.932 13	0.868 8	0.744 4
Peng-Shuai Wang, Yang Liu, Yu-Xiao Guo, Chun-Yu Sun, Xin Tong: O-CNN: Octree-based Convolutional Neural Networks for 3D Shape Analysis. SIGGRAPH 2017
SAT	0.742 15	0.860 13	0.765 32	0.819 16	0.769 3	0.848 12	0.533 10	0.829 18	0.663 12	0.631 15	0.955 17	0.586 8	0.274 31	0.753 13	0.896 6	0.729 14	0.760 34	0.666 14	0.921 16	0.855 18	0.733 7

PointTransformer++	0.725 18	0.727 57	0.811 11	0.819 16	0.765 4	0.841 15	0.502 20	0.814 25	0.621 23	0.623 17	0.955 17	0.556 15	0.284 27	0.620 35	0.866 14	0.781 5	0.757 37	0.648 17	0.932 13	0.862 12	0.709 17

PointTransformerV2	0.752 5	0.742 50	0.809 12	0.872 1	0.758 5	0.860 6	0.552 6	0.891 5	0.610 27	0.687 2	0.960 7	0.559 14	0.304 18	0.766 8	0.926 2	0.767 6	0.797 13	0.644 19	0.942 5	0.876 7	0.722 14
Xiaoyang Wu, Yixing Lao, Li Jiang, Xihui Liu, Hengshuang Zhao: Point Transformer V2: Grouped Vector Attention and Partition-based Pooling. NeurIPS 2022
CU-Hybrid Net	0.764 2	0.924 2	0.819 6	0.840 11	0.757 6	0.853 9	0.580 1	0.848 13	0.709 2	0.643 11	0.958 9	0.587 7	0.295 21	0.753 13	0.884 12	0.758 8	0.815 5	0.725 2	0.927 15	0.867 9	0.743 5

StratifiedFormer	0.747 10	0.901 6	0.803 14	0.845 8	0.757 6	0.846 13	0.512 17	0.825 19	0.696 4	0.645 9	0.956 13	0.576 9	0.262 41	0.744 17	0.861 16	0.742 12	0.770 29	0.705 5	0.899 29	0.860 13	0.734 6
Xin Lai, Jianhui Liu, Li Jiang, Liwei Wang, Hengshuang Zhao, Shu Liu, Xiaojuan Qi, Jiaya Jia: Stratified Transformer for 3D Point Cloud Segmentation. CVPR 2022
BPNet	0.749 7	0.909 4	0.818 8	0.811 21	0.752 8	0.839 17	0.485 29	0.842 15	0.673 9	0.644 10	0.957 12	0.528 23	0.305 17	0.773 6	0.859 17	0.788 4	0.818 4	0.693 7	0.916 19	0.856 16	0.723 13
Wenbo Hu, Hengshuang Zhao, Li Jiang, Jiaya Jia, Tien-Tsin Wong: Bidirectional Projection Network for Cross Dimension Scene Understanding. CVPR 2021 (Oral)
PointConvFormer	0.749 7	0.793 29	0.790 21	0.807 24	0.750 9	0.856 8	0.524 14	0.881 6	0.588 37	0.642 13	0.977 4	0.591 5	0.274 31	0.781 2	0.929 1	0.804 3	0.796 14	0.642 20	0.947 3	0.885 2	0.715 16
Wenxuan Wu, Qi Shan, Li Fuxin: PointConvFormer: Revenge of the Point-based Convolution.
OccuSeg+Semantic	0.764 2	0.758 43	0.796 17	0.839 12	0.746 10	0.907 1	0.562 4	0.850 12	0.680 8	0.672 5	0.978 2	0.610 1	0.335 8	0.777 4	0.819 30	0.847 1	0.830 1	0.691 8	0.972 1	0.885 2	0.727 10

MSP	0.748 9	0.623 74	0.804 13	0.859 3	0.745 11	0.824 30	0.501 21	0.912 2	0.690 5	0.685 3	0.956 13	0.567 11	0.320 13	0.768 7	0.918 3	0.720 19	0.802 9	0.676 10	0.921 16	0.881 4	0.779 1

Retro-FPN	0.744 13	0.842 18	0.800 15	0.767 38	0.740 12	0.836 21	0.541 8	0.914 1	0.672 10	0.626 16	0.958 9	0.552 17	0.272 33	0.777 4	0.886 11	0.696 29	0.801 10	0.674 11	0.941 6	0.858 14	0.717 15

MatchingNet	0.724 20	0.812 26	0.812 10	0.810 22	0.735 13	0.834 22	0.495 26	0.860 10	0.572 44	0.602 27	0.954 21	0.512 26	0.280 28	0.757 12	0.845 25	0.725 16	0.780 23	0.606 36	0.937 9	0.851 21	0.700 20

EQ-Net	0.743 14	0.620 75	0.799 16	0.849 5	0.730 14	0.822 32	0.493 27	0.897 4	0.664 11	0.681 4	0.955 17	0.562 13	0.378 1	0.760 11	0.903 4	0.738 13	0.801 10	0.673 12	0.907 22	0.877 5	0.745 3
Zetong Yang, Li Jiang, Yanan Sun, Bernt Schiele, Jiaya JIa: A Unified Query-based Paradigm for Point Cloud Understanding. CVPR 2022
VMNet	0.746 11	0.870 11	0.838 2	0.858 4	0.729 15	0.850 11	0.501 21	0.874 7	0.587 38	0.658 7	0.956 13	0.564 12	0.299 19	0.765 9	0.900 5	0.716 22	0.812 7	0.631 25	0.939 8	0.858 14	0.709 17
Zeyu HU, Xuyang Bai, Jiaxiang Shang, Runze Zhang, Jiayu Dong, Xin Wang, Guangyuan Sun, Hongbo Fu, Chiew-Lan Tai: VMNet: Voxel-Mesh Network for Geodesic-Aware 3D Semantic Segmentation. ICCV 2021 (Oral)
IPCA	0.731 17	0.890 7	0.837 3	0.864 2	0.726 16	0.873 2	0.530 13	0.824 20	0.489 68	0.647 8	0.978 2	0.609 2	0.336 7	0.624 34	0.733 43	0.758 8	0.776 25	0.570 50	0.949 2	0.877 5	0.728 8

SimConv	0.410 90	0.000 95	0.782 25	0.772 35	0.722 17	0.838 18	0.407 62	0.000 96	0.000 96	0.595 31	0.947 39	0.000 96	0.270 36	0.000 96	0.000 96	0.000 96	0.786 21	0.621 30	0.000 96	0.841 27	0.621 41

SparseConvNet	0.725 18	0.647 71	0.821 5	0.846 7	0.721 18	0.869 3	0.533 10	0.754 39	0.603 33	0.614 20	0.955 17	0.572 10	0.325 11	0.710 19	0.870 13	0.724 17	0.823 2	0.628 26	0.934 11	0.865 11	0.683 23

One Thing One Click	0.701 25	0.825 23	0.796 17	0.723 45	0.716 19	0.832 23	0.433 55	0.816 21	0.634 19	0.609 22	0.969 6	0.418 63	0.344 5	0.559 50	0.833 27	0.715 23	0.808 8	0.560 54	0.902 26	0.847 23	0.680 24

PicassoNet-II	0.696 27	0.704 61	0.790 21	0.787 30	0.709 20	0.837 19	0.459 38	0.815 23	0.543 53	0.615 19	0.956 13	0.529 21	0.250 44	0.551 55	0.790 35	0.703 27	0.799 12	0.619 31	0.908 21	0.848 22	0.700 20
Huan Lei, Naveed Akhtar, Mubarak Shah, and Ajmal Mian: Geometric feature learning for 3D meshes.
MinkowskiNet	0.736 16	0.859 14	0.818 8	0.832 13	0.709 20	0.840 16	0.521 16	0.853 11	0.660 14	0.643 11	0.951 28	0.544 18	0.286 26	0.731 18	0.893 7	0.675 36	0.772 27	0.683 9	0.874 47	0.852 20	0.727 10
C. Choy, J. Gwak, S. Savarese: 4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks. CVPR 2019
INS-Conv-semantic	0.717 21	0.751 46	0.759 35	0.812 20	0.704 22	0.868 4	0.537 9	0.842 15	0.609 29	0.608 23	0.953 23	0.534 19	0.293 22	0.616 36	0.864 15	0.719 21	0.793 18	0.640 21	0.933 12	0.845 25	0.663 28

Virtual MVFusion	0.746 11	0.771 37	0.819 6	0.848 6	0.702 23	0.865 5	0.397 66	0.899 3	0.699 3	0.664 6	0.948 37	0.588 6	0.330 9	0.746 16	0.851 23	0.764 7	0.796 14	0.704 6	0.935 10	0.866 10	0.728 8
Abhijit Kundu, Xiaoqi Yin, Alireza Fathi, David Ross, Brian Brewington, Thomas Funkhouser, Caroline Pantofaru: Virtual Multi-view Fusion for 3D Semantic Segmentation. ECCV 2020
VACNN++	0.684 32	0.728 56	0.757 38	0.776 34	0.690 24	0.804 49	0.464 36	0.816 21	0.577 43	0.587 34	0.945 46	0.508 28	0.276 30	0.671 21	0.710 48	0.663 41	0.750 40	0.589 45	0.881 41	0.832 29	0.653 31

DMF-Net	0.752 5	0.906 5	0.793 20	0.802 26	0.689 25	0.825 28	0.556 5	0.867 8	0.681 7	0.602 27	0.960 7	0.555 16	0.365 3	0.779 3	0.859 17	0.747 11	0.795 17	0.717 4	0.917 18	0.856 16	0.764 2

contrastBoundary	0.705 23	0.769 40	0.775 28	0.809 23	0.687 26	0.820 35	0.439 53	0.812 26	0.661 13	0.591 33	0.945 46	0.515 25	0.171 73	0.633 31	0.856 19	0.720 19	0.796 14	0.668 13	0.889 36	0.847 23	0.689 22
Liyao Tang, Yibing Zhan, Zhe Chen, Baosheng Yu, Dacheng Tao: Contrastive Boundary Learning for Point Cloud Segmentation. CVPR2022
One-Thing-One-Click	0.693 28	0.743 49	0.794 19	0.655 68	0.684 27	0.822 32	0.497 25	0.719 49	0.622 22	0.617 18	0.977 4	0.447 50	0.339 6	0.750 15	0.664 58	0.703 27	0.790 20	0.596 40	0.946 4	0.855 18	0.647 33
Zhengzhe Liu, Xiaojuan Qi, Chi-Wing Fu: One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentation. CVPR 2021
PointMetaBase	0.714 22	0.835 19	0.785 23	0.821 14	0.684 27	0.846 13	0.531 12	0.865 9	0.614 24	0.596 30	0.953 23	0.500 29	0.246 47	0.674 20	0.888 9	0.692 30	0.764 31	0.624 27	0.849 61	0.844 26	0.675 25

RFCR	0.702 24	0.889 8	0.745 44	0.813 19	0.672 29	0.818 39	0.493 27	0.815 23	0.623 21	0.610 21	0.947 39	0.470 38	0.249 46	0.594 39	0.848 24	0.705 26	0.779 24	0.646 18	0.892 34	0.823 33	0.611 43
Jingyu Gong, Jiachen Xu, Xin Tan, Haichuan Song, Yanyun Qu, Yuan Xie, Lizhuang Ma: Omni-Supervised Point Cloud Segmentation via Gradual Receptive Field Component Reasoning. CVPR2021
JSENet	0.699 26	0.881 10	0.762 33	0.821 14	0.667 30	0.800 51	0.522 15	0.792 31	0.613 25	0.607 24	0.935 65	0.492 31	0.205 60	0.576 44	0.853 21	0.691 31	0.758 36	0.652 16	0.872 50	0.828 30	0.649 32
Zeyu HU, Mingmin Zhen, Xuyang BAI, Hongbo Fu, Chiew-lan Tai: JSENet: Joint Semantic Segmentation and Edge Detection Network for 3D Point Clouds. ECCV 2020
FusionNet	0.688 30	0.704 61	0.741 48	0.754 42	0.656 31	0.829 25	0.501 21	0.741 44	0.609 29	0.548 40	0.950 32	0.522 24	0.371 2	0.633 31	0.756 38	0.715 23	0.771 28	0.623 28	0.861 57	0.814 38	0.658 29
Feihu Zhang, Jin Fang, Benjamin Wah, Philip Torr: Deep FusionNet for Point Cloud Semantic Segmentation. ECCV 2020
PointASNL	0.666 39	0.703 63	0.781 26	0.751 44	0.655 32	0.830 24	0.471 32	0.769 35	0.474 71	0.537 44	0.951 28	0.475 36	0.279 29	0.635 29	0.698 52	0.675 36	0.751 39	0.553 59	0.816 68	0.806 42	0.703 19
Xu Yan, Chaoda Zheng, Zhen Li, Sheng Wang, Shuguang Cui: PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Sampling. CVPR 2020
Superpoint Network	0.683 34	0.851 16	0.728 52	0.800 28	0.653 33	0.806 47	0.468 33	0.804 27	0.572 44	0.602 27	0.946 43	0.453 47	0.239 50	0.519 61	0.822 28	0.689 34	0.762 33	0.595 42	0.895 32	0.827 31	0.630 39

SALANet	0.670 38	0.816 25	0.770 30	0.768 37	0.652 34	0.807 46	0.451 40	0.747 41	0.659 15	0.545 41	0.924 75	0.473 37	0.149 83	0.571 46	0.811 32	0.635 54	0.746 41	0.623 28	0.892 34	0.794 50	0.570 60

KP-FCNN	0.684 32	0.847 17	0.758 37	0.784 32	0.647 35	0.814 42	0.473 31	0.772 34	0.605 31	0.594 32	0.935 65	0.450 48	0.181 71	0.587 40	0.805 33	0.690 32	0.785 22	0.614 32	0.882 40	0.819 37	0.632 38
H. Thomas, C. Qi, J. Deschaud, B. Marcotegui, F. Goulette, L. Guibas.: KPConv: Flexible and Deformable Convolution for Point Clouds. ICCV 2019
Feature_GeometricNet	0.690 29	0.884 9	0.754 39	0.795 29	0.647 35	0.818 39	0.422 57	0.802 29	0.612 26	0.604 25	0.945 46	0.462 41	0.189 68	0.563 49	0.853 21	0.726 15	0.765 30	0.632 24	0.904 24	0.821 36	0.606 47
Kangcheng Liu, Ben M. Chen: https://arxiv.org/abs/2012.09439. arXiv Preprint
Feature-Geometry Net	0.685 31	0.866 12	0.748 41	0.819 16	0.645 37	0.794 54	0.450 43	0.802 29	0.587 38	0.604 25	0.945 46	0.464 40	0.201 63	0.554 52	0.840 26	0.723 18	0.732 46	0.602 38	0.907 22	0.822 35	0.603 50

PointConv	0.666 39	0.781 32	0.759 35	0.699 53	0.644 38	0.822 32	0.475 30	0.779 32	0.564 49	0.504 58	0.953 23	0.428 57	0.203 62	0.586 42	0.754 39	0.661 42	0.753 38	0.588 46	0.902 26	0.813 40	0.642 34
Wenxuan Wu, Zhongang Qi, Li Fuxin: PointConv: Deep Convolutional Networks on 3D Point Clouds. CVPR 2019
PointContrast_LA_SEM	0.683 34	0.757 44	0.784 24	0.786 31	0.639 39	0.824 30	0.408 60	0.775 33	0.604 32	0.541 42	0.934 69	0.532 20	0.269 37	0.552 53	0.777 36	0.645 51	0.793 18	0.640 21	0.913 20	0.824 32	0.671 26

PPCNN++	0.663 41	0.746 47	0.708 55	0.722 46	0.638 40	0.820 35	0.451 40	0.566 75	0.599 35	0.541 42	0.950 32	0.510 27	0.313 14	0.648 26	0.819 30	0.616 59	0.682 64	0.590 44	0.869 53	0.810 41	0.656 30
Pyunghwan Ahn, Juyoung Yang, Eojindl Yi, Chanho Lee, Junmo Kim: Projection-based Point Convolution for Efficient Point Cloud Segmentation. IEEE Access
ROSMRF3D	0.673 37	0.789 30	0.748 41	0.763 40	0.635 41	0.814 42	0.407 62	0.747 41	0.581 42	0.573 35	0.950 32	0.484 32	0.271 35	0.607 37	0.754 39	0.649 46	0.774 26	0.596 40	0.883 39	0.823 33	0.606 47

joint point-based	0.634 54	0.614 76	0.778 27	0.667 65	0.633 42	0.825 28	0.420 58	0.804 27	0.467 73	0.561 37	0.951 28	0.494 30	0.291 23	0.566 47	0.458 73	0.579 70	0.764 31	0.559 56	0.838 63	0.814 38	0.598 52
Hung-Yueh Chiang, Yen-Liang Lin, Yueh-Cheng Liu, Winston H. Hsu: A Unified Point-Based Framework for 3D Segmentation. 3DV 2019
VI-PointConv	0.676 36	0.770 39	0.754 39	0.783 33	0.621 43	0.814 42	0.552 6	0.758 37	0.571 46	0.557 38	0.954 21	0.529 21	0.268 39	0.530 59	0.682 53	0.675 36	0.719 49	0.603 37	0.888 37	0.833 28	0.665 27
Xingyi Li, Wenxuan Wu, Xiaoli Z. Fern, Li Fuxin: The Devils in the Point Clouds: Studying the Robustness of Point Cloud Convolutions.
DCM-Net	0.658 42	0.778 33	0.702 58	0.806 25	0.619 44	0.813 45	0.468 33	0.693 57	0.494 64	0.524 50	0.941 57	0.449 49	0.298 20	0.510 63	0.821 29	0.675 36	0.727 48	0.568 52	0.826 66	0.803 44	0.637 36
Jonas Schult, Francis Engelmann, Theodora Kontogianni, Bastian Leibe: DualConvMesh-Net: Joint Geodesic and Euclidean Convolutions on 3D Meshes. CVPR 2020 [Oral]
Supervoxel-CNN	0.635 53	0.656 69	0.711 54	0.719 48	0.613 45	0.757 71	0.444 50	0.765 36	0.534 55	0.566 36	0.928 73	0.478 35	0.272 33	0.636 28	0.531 67	0.664 40	0.645 74	0.508 72	0.864 56	0.792 55	0.611 43

3DSM_DMMF	0.631 56	0.626 73	0.745 44	0.801 27	0.607 46	0.751 72	0.506 18	0.729 48	0.565 48	0.491 60	0.866 89	0.434 52	0.197 66	0.595 38	0.630 60	0.709 25	0.705 56	0.560 54	0.875 45	0.740 74	0.491 78

FPConv	0.639 49	0.785 31	0.760 34	0.713 51	0.603 47	0.798 52	0.392 68	0.534 80	0.603 33	0.524 50	0.948 37	0.457 43	0.250 44	0.538 57	0.723 46	0.598 64	0.696 59	0.614 32	0.872 50	0.799 45	0.567 62
Yiqun Lin, Zizheng Yan, Haibin Huang, Dong Du, Ligang Liu, Shuguang Cui, Xiaoguang Han: FPConv: Learning Local Flattening for Point Convolution. CVPR 2020
SegGroup_sem	0.627 61	0.818 24	0.747 43	0.701 52	0.602 48	0.764 68	0.385 72	0.629 67	0.490 66	0.508 55	0.931 72	0.409 65	0.201 63	0.564 48	0.725 45	0.618 57	0.692 60	0.539 66	0.873 48	0.794 50	0.548 69
An Tao, Yueqi Duan, Yi Wei, Jiwen Lu, Jie Zhou: SegGroup: Seg-Level Supervision for 3D Instance and Semantic Segmentation. TIP 2022
wsss-transformer	0.600 66	0.634 72	0.743 46	0.697 55	0.601 49	0.781 60	0.437 54	0.585 73	0.493 65	0.446 69	0.933 70	0.394 68	0.011 94	0.654 24	0.661 59	0.603 61	0.733 45	0.526 69	0.832 64	0.761 69	0.480 80

HPEIN	0.618 63	0.729 55	0.668 72	0.647 71	0.597 50	0.766 67	0.414 59	0.680 58	0.520 58	0.525 49	0.946 43	0.432 53	0.215 56	0.493 68	0.599 62	0.638 52	0.617 79	0.570 50	0.897 30	0.806 42	0.605 49
Li Jiang, Hengshuang Zhao, Shu Liu, Xiaoyong Shen, Chi-Wing Fu, Jiaya Jia: Hierarchical Point-Edge Interaction Network for Point Cloud Semantic Segmentation. ICCV 2019
PointMTL	0.632 55	0.731 54	0.688 69	0.675 60	0.591 51	0.784 59	0.444 50	0.565 76	0.610 27	0.492 59	0.949 35	0.456 44	0.254 43	0.587 40	0.706 49	0.599 63	0.665 70	0.612 35	0.868 54	0.791 58	0.579 57

FusionAwareConv	0.630 59	0.604 78	0.741 48	0.766 39	0.590 52	0.747 73	0.501 21	0.734 46	0.503 63	0.527 48	0.919 79	0.454 45	0.323 12	0.550 56	0.420 77	0.678 35	0.688 62	0.544 62	0.896 31	0.795 49	0.627 40
Jiazhao Zhang, Chenyang Zhu, Lintao Zheng, Kai Xu: Fusion-Aware Point Convolution for Online Semantic 3D Scene Segmentation. CVPR 2020
PD-Net	0.638 50	0.797 28	0.769 31	0.641 74	0.590 52	0.820 35	0.461 37	0.537 79	0.637 18	0.536 45	0.947 39	0.388 70	0.206 59	0.656 23	0.668 56	0.647 49	0.732 46	0.585 47	0.868 54	0.793 52	0.473 83

MVPNet	0.641 46	0.831 20	0.715 53	0.671 63	0.590 52	0.781 60	0.394 67	0.679 59	0.642 16	0.553 39	0.937 62	0.462 41	0.256 42	0.649 25	0.406 78	0.626 55	0.691 61	0.666 14	0.877 43	0.792 55	0.608 46
Maximilian Jaritz, Jiayuan Gu, Hao Su: Multi-view PointNet for 3D Scene Understanding. GMDL Workshop, ICCV 2019
DVVNet	0.562 74	0.648 70	0.700 60	0.770 36	0.586 55	0.687 80	0.333 79	0.650 63	0.514 61	0.475 64	0.906 83	0.359 74	0.223 53	0.340 81	0.442 76	0.422 85	0.668 69	0.501 73	0.708 80	0.779 61	0.534 72

SAFNet-seg	0.654 44	0.752 45	0.734 50	0.664 66	0.583 56	0.815 41	0.399 65	0.754 39	0.639 17	0.535 46	0.942 55	0.470 38	0.309 16	0.665 22	0.539 65	0.650 45	0.708 54	0.635 23	0.857 59	0.793 52	0.642 34
Linqing Zhao, Jiwen Lu, Jie Zhou: Similarity-Aware Fusion Network for 3D Semantic Segmentation. IROS 2021
RandLA-Net	0.645 45	0.778 33	0.731 51	0.699 53	0.577 57	0.829 25	0.446 45	0.736 45	0.477 70	0.523 52	0.945 46	0.454 45	0.269 37	0.484 70	0.749 42	0.618 57	0.738 42	0.599 39	0.827 65	0.792 55	0.621 41

PointSPNet	0.637 51	0.734 53	0.692 66	0.714 50	0.576 58	0.797 53	0.446 45	0.743 43	0.598 36	0.437 72	0.942 55	0.403 66	0.150 82	0.626 33	0.800 34	0.649 46	0.697 58	0.557 57	0.846 62	0.777 63	0.563 63

PointMRNet	0.640 48	0.717 60	0.701 59	0.692 56	0.576 58	0.801 50	0.467 35	0.716 50	0.563 50	0.459 67	0.953 23	0.429 56	0.169 75	0.581 43	0.854 20	0.605 60	0.710 51	0.550 60	0.894 33	0.793 52	0.575 58

SConv	0.636 52	0.830 21	0.697 62	0.752 43	0.572 60	0.780 62	0.445 47	0.716 50	0.529 56	0.530 47	0.951 28	0.446 51	0.170 74	0.507 65	0.666 57	0.636 53	0.682 64	0.541 65	0.886 38	0.799 45	0.594 54

HPGCNN	0.656 43	0.698 64	0.743 46	0.650 69	0.564 61	0.820 35	0.505 19	0.758 37	0.631 20	0.479 62	0.945 46	0.480 34	0.226 51	0.572 45	0.774 37	0.690 32	0.735 44	0.614 32	0.853 60	0.776 64	0.597 53
Jisheng Dang, Qingyong Hu, Yulan Guo, Jun Yang: HPGCNN.
ROSMRF	0.580 70	0.772 36	0.707 56	0.681 59	0.563 62	0.764 68	0.362 75	0.515 81	0.465 74	0.465 66	0.936 64	0.427 59	0.207 58	0.438 73	0.577 63	0.536 74	0.675 67	0.486 77	0.723 79	0.779 61	0.524 74

SIConv	0.625 62	0.830 21	0.694 64	0.757 41	0.563 62	0.772 66	0.448 44	0.647 65	0.520 58	0.509 54	0.949 35	0.431 55	0.191 67	0.496 67	0.614 61	0.647 49	0.672 68	0.535 68	0.876 44	0.783 60	0.571 59

PointConv-SFPN	0.641 46	0.776 35	0.703 57	0.721 47	0.557 64	0.826 27	0.451 40	0.672 61	0.563 50	0.483 61	0.943 54	0.425 60	0.162 78	0.644 27	0.726 44	0.659 43	0.709 53	0.572 49	0.875 45	0.786 59	0.559 65

APCF-Net	0.631 56	0.742 50	0.687 71	0.672 61	0.557 64	0.792 57	0.408 60	0.665 62	0.545 52	0.508 55	0.952 27	0.428 57	0.186 69	0.634 30	0.702 50	0.620 56	0.706 55	0.555 58	0.873 48	0.798 47	0.581 56
Haojia, Lin: Adaptive Pyramid Context Fusion for Point Cloud Perception. GRSL
DenSeR	0.628 60	0.800 27	0.625 81	0.719 48	0.545 66	0.806 47	0.445 47	0.597 70	0.448 77	0.519 53	0.938 61	0.481 33	0.328 10	0.489 69	0.499 72	0.657 44	0.759 35	0.592 43	0.881 41	0.797 48	0.634 37

SPH3D-GCN	0.610 64	0.858 15	0.772 29	0.489 86	0.532 67	0.792 57	0.404 64	0.643 66	0.570 47	0.507 57	0.935 65	0.414 64	0.046 92	0.510 63	0.702 50	0.602 62	0.705 56	0.549 61	0.859 58	0.773 65	0.534 72
Huan Lei, Naveed Akhtar, and Ajmal Mian: Spherical Kernel for Efficient Graph Convolution on 3D Point Clouds. TPAMI 2020
PointNet2-SFPN	0.631 56	0.771 37	0.692 66	0.672 61	0.524 68	0.837 19	0.440 52	0.706 55	0.538 54	0.446 69	0.944 52	0.421 62	0.219 54	0.552 53	0.751 41	0.591 66	0.737 43	0.543 64	0.901 28	0.768 66	0.557 66

AttAN	0.609 65	0.760 42	0.667 73	0.649 70	0.521 69	0.793 55	0.457 39	0.648 64	0.528 57	0.434 74	0.947 39	0.401 67	0.153 81	0.454 72	0.721 47	0.648 48	0.717 50	0.536 67	0.904 24	0.765 67	0.485 79
Gege Zhang, Qinghua Ma, Licheng Jiao, Fang Liu and Qigong Sun: AttAN: Attention Adversarial Networks for 3D Point Cloud Semantic Segmentation. IJCAI2020
SD-DETR	0.576 71	0.746 47	0.609 85	0.445 90	0.517 70	0.643 86	0.366 74	0.714 52	0.456 75	0.468 65	0.870 88	0.432 53	0.264 40	0.558 51	0.674 54	0.586 69	0.688 62	0.482 78	0.739 77	0.733 76	0.537 71

SQN_0.1%	0.569 72	0.676 66	0.696 63	0.657 67	0.497 71	0.779 63	0.424 56	0.548 77	0.515 60	0.376 79	0.902 86	0.422 61	0.357 4	0.379 79	0.456 74	0.596 65	0.659 71	0.544 62	0.685 82	0.665 87	0.556 67

TextureNet	0.566 73	0.672 68	0.664 74	0.671 63	0.494 72	0.719 76	0.445 47	0.678 60	0.411 83	0.396 77	0.935 65	0.356 75	0.225 52	0.412 77	0.535 66	0.565 72	0.636 78	0.464 80	0.794 71	0.680 84	0.568 61
Jingwei Huang, Haotian Zhang, Li Yi, Thomas Funkerhouser, Matthias Niessner, Leonidas Guibas: TextureNet: Consistent Local Parametrizations for Learning from High-Resolution Signals on Meshes. CVPR
Pointnet++ & Feature	0.557 75	0.735 52	0.661 75	0.686 57	0.491 73	0.744 74	0.392 68	0.539 78	0.451 76	0.375 80	0.946 43	0.376 72	0.205 60	0.403 78	0.356 81	0.553 73	0.643 75	0.497 74	0.824 67	0.756 70	0.515 75

DPC	0.592 68	0.720 58	0.700 60	0.602 79	0.480 74	0.762 70	0.380 73	0.713 53	0.585 41	0.437 72	0.940 59	0.369 73	0.288 24	0.434 75	0.509 71	0.590 68	0.639 77	0.567 53	0.772 73	0.755 71	0.592 55
Francis Engelmann, Theodora Kontogianni, Bastian Leibe: Dilated Point Convolutions: On the Receptive Field Size of Point Convolutions on 3D Point Clouds. ICRA 2020
3DMV, FTSDF	0.501 80	0.558 82	0.608 86	0.424 92	0.478 75	0.690 79	0.246 88	0.586 72	0.468 72	0.450 68	0.911 81	0.394 68	0.160 79	0.438 73	0.212 88	0.432 84	0.541 87	0.475 79	0.742 76	0.727 77	0.477 81

GMLPs	0.538 76	0.495 86	0.693 65	0.647 71	0.471 76	0.793 55	0.300 82	0.477 82	0.505 62	0.358 81	0.903 85	0.327 79	0.081 89	0.472 71	0.529 68	0.448 83	0.710 51	0.509 70	0.746 75	0.737 75	0.554 68

CCRFNet	0.589 69	0.766 41	0.659 76	0.683 58	0.470 77	0.740 75	0.387 71	0.620 69	0.490 66	0.476 63	0.922 77	0.355 76	0.245 48	0.511 62	0.511 70	0.571 71	0.643 75	0.493 76	0.872 50	0.762 68	0.600 51

LAP-D	0.594 67	0.720 58	0.692 66	0.637 75	0.456 78	0.773 65	0.391 70	0.730 47	0.587 38	0.445 71	0.940 59	0.381 71	0.288 24	0.434 75	0.453 75	0.591 66	0.649 72	0.581 48	0.777 72	0.749 73	0.610 45

subcloud_weak	0.516 78	0.676 66	0.591 88	0.609 76	0.442 79	0.774 64	0.335 78	0.597 70	0.422 82	0.357 82	0.932 71	0.341 78	0.094 88	0.298 83	0.528 69	0.473 81	0.676 66	0.495 75	0.602 88	0.721 79	0.349 90

Online SegFusion	0.515 79	0.607 77	0.644 79	0.579 81	0.434 80	0.630 88	0.353 76	0.628 68	0.440 78	0.410 75	0.762 93	0.307 81	0.167 76	0.520 60	0.403 79	0.516 75	0.565 82	0.447 84	0.678 83	0.701 81	0.514 76
Davide Menini, Suryansh Kumar, Martin R. Oswald, Erik Sandstroem, Cristian Sminchisescu, Luc van Gool: A Real-Time Learning Framework for Joint 3D Reconstruction and Semantic Segmentation. Robotics and Automation Letters Submission
3DMV	0.484 82	0.484 88	0.538 90	0.643 73	0.424 81	0.606 91	0.310 80	0.574 74	0.433 81	0.378 78	0.796 91	0.301 82	0.214 57	0.537 58	0.208 89	0.472 82	0.507 91	0.413 89	0.693 81	0.602 90	0.539 70
Angela Dai, Matthias Niessner: 3DMV: Joint 3D-Multi-View Prediction for 3D Semantic Scene Segmentation. ECCV'18
PCNN	0.498 81	0.559 81	0.644 79	0.560 83	0.420 82	0.711 78	0.229 90	0.414 83	0.436 79	0.352 83	0.941 57	0.324 80	0.155 80	0.238 88	0.387 80	0.493 77	0.529 88	0.509 70	0.813 69	0.751 72	0.504 77

PanopticFusion-label	0.529 77	0.491 87	0.688 69	0.604 78	0.386 83	0.632 87	0.225 92	0.705 56	0.434 80	0.293 87	0.815 90	0.348 77	0.241 49	0.499 66	0.669 55	0.507 76	0.649 72	0.442 86	0.796 70	0.602 90	0.561 64
Gaku Narita, Takashi Seno, Tomoya Ishikawa, Yohsuke Kaji: PanopticFusion: Online Volumetric Semantic Mapping at the Level of Stuff and Things. IROS 2019 (to appear)
FCPN	0.447 84	0.679 65	0.604 87	0.578 82	0.380 84	0.682 81	0.291 85	0.106 93	0.483 69	0.258 92	0.920 78	0.258 88	0.025 93	0.231 90	0.325 82	0.480 80	0.560 84	0.463 81	0.725 78	0.666 86	0.231 94
Dario Rethage, Johanna Wald, Jürgen Sturm, Nassir Navab, Federico Tombari: Fully-Convolutional Point Networks for Large-Scale Point Clouds. ECCV 2018
Tangent Convolutions	0.438 88	0.437 91	0.646 78	0.474 87	0.369 85	0.645 85	0.353 76	0.258 90	0.282 92	0.279 88	0.918 80	0.298 83	0.147 84	0.283 85	0.294 83	0.487 78	0.562 83	0.427 88	0.619 87	0.633 88	0.352 89
Maxim Tatarchenko, Jaesik Park, Vladlen Koltun, Qian-Yi Zhou: Tangent convolutions for dense prediction in 3d. CVPR 2018
DGCNN_reproduce	0.446 85	0.474 89	0.623 82	0.463 88	0.366 86	0.651 84	0.310 80	0.389 86	0.349 88	0.330 84	0.937 62	0.271 86	0.126 85	0.285 84	0.224 87	0.350 90	0.577 81	0.445 85	0.625 86	0.723 78	0.394 86
Yue Wang, Yongbin Sun, Ziwei Liu, Sanjay E. Sarma, Michael M. Bronstein, Justin M. Solomon: Dynamic Graph CNN for Learning on Point Clouds. TOG 2019
PNET2	0.442 86	0.548 83	0.548 89	0.597 80	0.363 87	0.628 89	0.300 82	0.292 88	0.374 85	0.307 86	0.881 87	0.268 87	0.186 69	0.238 88	0.204 90	0.407 86	0.506 92	0.449 83	0.667 84	0.620 89	0.462 84

ScanNet+FTSDF	0.383 92	0.297 93	0.491 92	0.432 91	0.358 88	0.612 90	0.274 86	0.116 92	0.411 83	0.265 90	0.904 84	0.229 90	0.079 90	0.250 86	0.185 91	0.320 91	0.510 89	0.385 91	0.548 90	0.597 93	0.394 86

SurfaceConvPF	0.442 86	0.505 85	0.622 83	0.380 93	0.342 89	0.654 83	0.227 91	0.397 85	0.367 86	0.276 89	0.924 75	0.240 89	0.198 65	0.359 80	0.262 84	0.366 87	0.581 80	0.435 87	0.640 85	0.668 85	0.398 85
Hao Pan, Shilin Liu, Yang Liu, Xin Tong: Convolutional Neural Networks on 3D Surfaces Using Parallel Frames.
3DWSSS	0.425 89	0.525 84	0.647 77	0.522 84	0.324 90	0.488 94	0.077 95	0.712 54	0.353 87	0.401 76	0.636 95	0.281 85	0.176 72	0.340 81	0.565 64	0.175 94	0.551 85	0.398 90	0.370 94	0.602 90	0.361 88

PointCNN with RGB	0.458 83	0.577 80	0.611 84	0.356 94	0.321 91	0.715 77	0.299 84	0.376 87	0.328 90	0.319 85	0.944 52	0.285 84	0.164 77	0.216 91	0.229 86	0.484 79	0.545 86	0.456 82	0.755 74	0.709 80	0.475 82
Yangyan Li, Rui Bu, Mingchao Sun, Baoquan Chen: PointCNN. NeurIPS 2018
ScanNet	0.306 95	0.203 94	0.366 94	0.501 85	0.311 92	0.524 93	0.211 93	0.002 95	0.342 89	0.189 94	0.786 92	0.145 94	0.102 87	0.245 87	0.152 92	0.318 92	0.348 94	0.300 94	0.460 93	0.437 95	0.182 95
Angela Dai, Angel X. Chang, Manolis Savva, Maciej Halber, Thomas Funkhouser, Matthias Nießner: ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes. CVPR'17
SPLAT Net	0.393 91	0.472 90	0.511 91	0.606 77	0.311 92	0.656 82	0.245 89	0.405 84	0.328 90	0.197 93	0.927 74	0.227 91	0.000 96	0.001 95	0.249 85	0.271 93	0.510 89	0.383 92	0.593 89	0.699 82	0.267 92
Hang Su, Varun Jampani, Deqing Sun, Subhransu Maji, Evangelos Kalogerakis, Ming-Hsuan Yang, Jan Kautz: SPLATNet: Sparse Lattice Networks for Point Cloud Processing. CVPR 2018
PointNet++	0.339 93	0.584 79	0.478 93	0.458 89	0.256 94	0.360 95	0.250 87	0.247 91	0.278 93	0.261 91	0.677 94	0.183 92	0.117 86	0.212 92	0.145 93	0.364 88	0.346 95	0.232 95	0.548 90	0.523 94	0.252 93
Charles R. Qi, Li Yi, Hao Su, Leonidas J. Guibas: pointnet++: deep hierarchical feature learning on point sets in a metric space.
SSC-UNet	0.308 94	0.353 92	0.290 95	0.278 95	0.166 95	0.553 92	0.169 94	0.286 89	0.147 94	0.148 95	0.908 82	0.182 93	0.064 91	0.023 94	0.018 95	0.354 89	0.363 93	0.345 93	0.546 92	0.685 83	0.278 91

ERROR	0.054 96	0.000 95	0.041 96	0.172 96	0.030 96	0.062 96	0.001 96	0.035 94	0.004 95	0.051 96	0.143 96	0.019 95	0.003 95	0.041 93	0.050 94	0.003 95	0.054 96	0.018 96	0.005 95	0.264 96	0.082 96

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

Method	avg ap 50%	bathtub	bed	bookshelf	cabinet	chair	counter	curtain	desk	door	otherfurniture	picture	refrigerator	shower curtain	sink	sofa	table	toilet	window

SoftGroup	0.761 4	1.000 1	0.808 15	0.845 6	0.716 1	0.862 7	0.243 14	0.824 3	0.655 8	0.620 4	0.734 2	0.699 5	0.791 4	0.981 22	0.716 6	0.844 4	0.769 5	1.000 1	0.594 7
Thang Vu, Kookhoi Kim, Tung M. Luu, Xuan Thanh Nguyen, Chang D. Yoo: SoftGroup for 3D Instance Segmentaiton on Point Clouds. CVPR 2022 [Oral]
PBNet	0.747 5	1.000 1	0.818 11	0.837 8	0.713 2	0.844 9	0.457 4	0.647 23	0.711 3	0.614 5	0.617 9	0.657 8	0.650 10	1.000 1	0.692 9	0.822 8	0.765 6	1.000 1	0.595 6
W.Zhao, Y.Yan, C.Yang, J.Ye,X.Yang,K.Huang: Divide and Conquer: 3D Instance Segmentation With Point-Wise Binarization. Arxiv
IPCA-Inst	0.731 7	1.000 1	0.788 23	0.884 5	0.698 3	0.788 23	0.252 12	0.760 12	0.646 9	0.511 15	0.637 8	0.665 7	0.804 3	1.000 1	0.644 16	0.778 13	0.747 8	1.000 1	0.561 13

Mask3D	0.780 1	1.000 1	0.786 24	0.716 22	0.696 4	0.885 3	0.500 2	0.714 17	0.810 1	0.672 3	0.715 3	0.679 6	0.809 1	1.000 1	0.831 1	0.833 7	0.787 3	1.000 1	0.602 4
Jonas Schult, Francis Engelmann, Alexander Hermans, Or Litany, Siyu Tang, Bastian Leibe: Mask3D for 3D Semantic Instance Segmentation. ICRA 2023
SoftGroup++	0.769 3	1.000 1	0.803 18	0.937 1	0.684 5	0.865 5	0.213 17	0.870 2	0.664 6	0.571 7	0.758 1	0.702 4	0.807 2	1.000 1	0.653 15	0.902 1	0.792 2	1.000 1	0.626 1

HAIS	0.699 10	1.000 1	0.849 5	0.820 9	0.675 6	0.808 16	0.279 8	0.757 13	0.465 19	0.517 14	0.596 10	0.559 12	0.600 20	1.000 1	0.654 14	0.767 15	0.676 14	0.994 31	0.560 14
Shaoyu Chen, Jiemin Fang, Qian Zhang, Wenyu Liu, Xinggang Wang: Hierarchical Aggregation for 3D Instance Segmentation. ICCV 2021
TopoSeg	0.725 8	1.000 1	0.806 17	0.933 2	0.668 7	0.758 26	0.272 10	0.734 16	0.630 10	0.549 11	0.654 5	0.606 10	0.697 7	0.966 24	0.612 20	0.839 5	0.754 7	1.000 1	0.573 9

GraphCut	0.732 6	1.000 1	0.788 22	0.724 21	0.642 8	0.859 8	0.248 13	0.787 10	0.618 12	0.596 6	0.653 6	0.722 2	0.583 26	1.000 1	0.766 4	0.861 2	0.825 1	1.000 1	0.504 19

DKNet	0.718 9	1.000 1	0.814 12	0.782 13	0.619 9	0.872 4	0.224 15	0.751 14	0.569 14	0.677 2	0.585 12	0.724 1	0.633 18	0.981 22	0.515 27	0.819 9	0.736 9	1.000 1	0.617 2
Yizheng Wu, Min Shi, Shuaiyuan Du, Hao Lu, Zhiguo Cao, Weicai Zhong: 3D Instances as 1D Kernels. ECCV 2022
SphereSeg	0.680 13	1.000 1	0.856 4	0.744 20	0.618 10	0.893 1	0.151 19	0.651 22	0.713 2	0.537 12	0.579 15	0.430 24	0.651 9	1.000 1	0.389 36	0.744 22	0.697 10	0.991 32	0.601 5

Mask-Group	0.664 16	1.000 1	0.822 10	0.764 18	0.616 11	0.815 13	0.139 23	0.694 19	0.597 13	0.459 21	0.566 16	0.599 11	0.600 20	0.516 43	0.715 7	0.819 10	0.635 19	1.000 1	0.603 3
Min Zhong, Xinghao Chen, Xiaokang Chen, Gang Zeng, Yunhe Wang: MaskGroup: Hierarchical Point Grouping and Masking for 3D Instance Segmentation. ICME 2022
SPFormer	0.770 2	0.903 34	0.903 1	0.806 10	0.609 12	0.886 2	0.568 1	0.815 6	0.705 4	0.711 1	0.655 4	0.652 9	0.685 8	1.000 1	0.789 3	0.809 11	0.776 4	1.000 1	0.583 8
Sun Jiahao, Qing Chunmei, Tan Junpeng, Xu Xiangmin: Superpoint Transformer for 3D Scene Instance Segmentation. AAAI 2023
SSEC	0.699 10	1.000 1	0.848 6	0.763 19	0.609 13	0.792 21	0.262 11	0.824 3	0.627 11	0.535 13	0.547 20	0.481 16	0.600 20	1.000 1	0.712 8	0.731 25	0.689 13	1.000 1	0.563 12

DENet	0.629 25	1.000 1	0.797 20	0.608 29	0.589 14	0.627 36	0.219 16	0.882 1	0.310 31	0.402 33	0.383 36	0.396 28	0.650 10	1.000 1	0.663 12	0.543 44	0.691 12	1.000 1	0.568 11

INS-Conv-instance	0.657 17	1.000 1	0.760 29	0.667 26	0.581 15	0.863 6	0.323 6	0.655 21	0.477 17	0.473 19	0.549 18	0.432 23	0.650 10	1.000 1	0.655 13	0.738 23	0.585 25	0.944 36	0.472 26

RWSeg	0.567 30	0.528 45	0.708 38	0.626 27	0.580 16	0.745 29	0.063 32	0.627 24	0.240 35	0.400 34	0.497 24	0.464 18	0.515 29	1.000 1	0.475 29	0.745 21	0.571 27	1.000 1	0.429 30

OccuSeg+instance	0.672 15	1.000 1	0.758 31	0.682 24	0.576 17	0.842 10	0.477 3	0.504 36	0.524 15	0.567 8	0.585 14	0.451 19	0.557 27	1.000 1	0.751 5	0.797 12	0.563 29	1.000 1	0.467 27
Lei Han, Tian Zheng, Lan Xu, Lu Fang: OccuSeg: Occupancy-aware 3D Instance Segmentation. CVPR2020
DD-UNet+Group	0.635 24	0.667 35	0.797 21	0.714 23	0.562 18	0.774 25	0.146 20	0.810 8	0.429 22	0.476 18	0.546 21	0.399 27	0.633 18	1.000 1	0.632 17	0.722 27	0.609 21	1.000 1	0.514 16
H. Liu, R. Liu, K. Yang, J. Zhang, K. Peng, R. Stiefelhagen: HIDA: Towards Holistic Indoor Understanding for the Visually Impaired via Semantic Instance Segmentation with a Wearable Solid-State LiDAR Sensor. ICCVW 2021
SSTNet	0.698 12	1.000 1	0.697 39	0.888 4	0.556 19	0.803 17	0.387 5	0.626 25	0.417 23	0.556 10	0.585 13	0.702 3	0.600 20	1.000 1	0.824 2	0.720 28	0.692 11	1.000 1	0.509 18
Zhihao Liang, Zhihao Li, Songcen Xu, Mingkui Tan, Kui Jia: Instance Segmentation in 3D Scenes using Semantic Superpoint Tree Networks. ICCV2021
RPGN	0.643 20	1.000 1	0.758 30	0.582 35	0.539 20	0.826 11	0.046 34	0.765 11	0.372 27	0.436 27	0.588 11	0.539 15	0.650 10	1.000 1	0.577 21	0.750 20	0.653 17	0.997 26	0.495 22
Shichao Dong, Guosheng Lin, Tzu-Yi Hung: Learning Regional Purity for Instance Segmentation on 3D Point Clouds. ECCV 2022
PE	0.645 19	1.000 1	0.773 26	0.798 12	0.538 21	0.786 24	0.088 30	0.799 9	0.350 29	0.435 28	0.547 19	0.545 13	0.646 17	0.933 25	0.562 23	0.761 18	0.556 34	0.997 26	0.501 21
Biao Zhang, Peter Wonka: Point Cloud Instance Segmentation using Probabilistic Embeddings. CVPR 2021
OSIS	0.605 27	1.000 1	0.801 19	0.599 30	0.535 22	0.728 30	0.286 7	0.436 40	0.679 5	0.491 16	0.433 30	0.256 33	0.404 40	0.857 27	0.620 19	0.724 26	0.510 38	1.000 1	0.539 15

Dyco3D	0.641 21	1.000 1	0.841 8	0.893 3	0.531 23	0.802 18	0.115 27	0.588 30	0.448 20	0.438 25	0.537 22	0.430 25	0.550 28	0.857 27	0.534 25	0.764 17	0.657 15	0.987 33	0.568 10
Tong He; Chunhua Shen; Anton van den Hengel: DyCo3D: Robust Instance Segmentation of 3D Point Clouds through Dynamic Convolution. CVPR2021
3D-MPA	0.611 26	1.000 1	0.833 9	0.765 17	0.526 24	0.756 27	0.136 25	0.588 30	0.470 18	0.438 26	0.432 32	0.358 31	0.650 10	0.857 27	0.429 32	0.765 16	0.557 32	1.000 1	0.430 29
Francis Engelmann, Martin Bokeloh, Alireza Fathi, Bastian Leibe, Matthias Nießner: 3D-MPA: Multi Proposal Aggregation for 3D Semantic Instance Segmentation. CVPR 2020
CSC-Pretrained	0.648 18	1.000 1	0.810 13	0.768 16	0.523 25	0.813 14	0.143 22	0.819 5	0.389 25	0.422 29	0.511 23	0.443 21	0.650 10	1.000 1	0.624 18	0.732 24	0.634 20	1.000 1	0.375 33

NeuralBF	0.555 31	0.667 35	0.896 2	0.843 7	0.517 26	0.751 28	0.029 36	0.519 34	0.414 24	0.439 24	0.465 26	0.000 51	0.484 31	0.857 27	0.287 40	0.693 33	0.651 18	1.000 1	0.485 24
Weiwei Sun, Daniel Rebain, Renjie Liao, Vladimir Tankovich, Soroosh Yazdani, Kwang Moo Yi, Andrea Tagliasacchi: NeuralBF: Neural Bilateral Filtering for Top-down Instance Segmentation on Point Clouds. WACV 2023
Box2Mask	0.677 14	1.000 1	0.847 7	0.771 15	0.509 27	0.816 12	0.277 9	0.558 32	0.482 16	0.562 9	0.640 7	0.448 20	0.700 5	1.000 1	0.666 10	0.852 3	0.578 26	0.997 26	0.488 23
Julian Chibane, Francis Engelmann, Tuan Anh Tran, Gerard Pons-Moll: Box2Mask: Weakly Supervised 3D Semantic Instance Segmentation Using Bounding Boxes. ECCV 2022
Occipital-SCS	0.512 35	1.000 1	0.716 35	0.509 36	0.506 28	0.611 37	0.092 29	0.602 29	0.177 38	0.346 38	0.383 35	0.165 38	0.442 35	0.850 34	0.386 37	0.618 40	0.543 35	0.889 40	0.389 32

PointGroup	0.636 23	1.000 1	0.765 27	0.624 28	0.505 29	0.797 19	0.116 26	0.696 18	0.384 26	0.441 23	0.559 17	0.476 17	0.596 24	1.000 1	0.666 10	0.756 19	0.556 33	0.997 26	0.513 17
Li Jiang, Hengshuang Zhao, Shaoshuai Shi, Shu Liu, Chi-Wing Fu, Jiaya Jia: PointGroup: Dual-Set Point Grouping for 3D Instance Segmentation. CVPR 2020 [oral]
GICN	0.638 22	1.000 1	0.895 3	0.800 11	0.480 30	0.676 32	0.144 21	0.737 15	0.354 28	0.447 22	0.400 34	0.365 30	0.700 5	1.000 1	0.569 22	0.836 6	0.599 22	1.000 1	0.473 25

SSEN	0.575 29	1.000 1	0.761 28	0.473 37	0.477 31	0.795 20	0.066 31	0.529 33	0.658 7	0.460 20	0.461 28	0.380 29	0.331 42	0.859 26	0.401 35	0.692 34	0.653 16	1.000 1	0.348 35
Dongsu Zhang, Junha Chun, Sang Kyun Cha, Young Min Kim: Spatial Semantic Embedding Network: Fast 3D Instance Segmentation with Deep Metric Learning. Arxiv
Sparse R-CNN	0.515 34	1.000 1	0.538 46	0.282 40	0.468 32	0.790 22	0.173 18	0.345 42	0.429 21	0.413 32	0.484 25	0.176 37	0.595 25	0.591 41	0.522 26	0.668 36	0.476 39	0.986 34	0.327 36

PCJC	0.578 28	1.000 1	0.810 14	0.583 34	0.449 33	0.813 15	0.042 35	0.603 28	0.341 30	0.490 17	0.465 27	0.410 26	0.650 10	0.835 35	0.264 42	0.694 32	0.561 30	0.889 40	0.504 20

One_Thing_One_Click	0.529 33	0.667 35	0.718 34	0.777 14	0.399 34	0.683 31	0.000 44	0.669 20	0.138 40	0.391 35	0.374 37	0.539 14	0.360 41	0.641 40	0.556 24	0.774 14	0.593 23	0.997 26	0.251 40
Zhengzhe Liu, Xiaojuan Qi, Chi-Wing Fu: One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentation. CVPR 2021
MASC	0.447 40	0.528 45	0.555 44	0.381 38	0.382 35	0.633 35	0.002 42	0.509 35	0.260 33	0.361 37	0.432 31	0.327 32	0.451 33	0.571 42	0.367 38	0.639 38	0.386 40	0.980 35	0.276 39
Chen Liu, Yasutaka Furukawa: MASC: Multi-scale Affinity with Sparse Convolution for 3D Instance Segmentation.
SPG_WSIS	0.470 38	0.667 35	0.685 40	0.677 25	0.372 36	0.562 41	0.000 44	0.482 37	0.244 34	0.316 41	0.298 39	0.052 46	0.442 36	0.857 27	0.267 41	0.702 29	0.559 31	1.000 1	0.287 38

R-PointNet	0.306 45	0.500 47	0.405 49	0.311 39	0.348 37	0.589 38	0.054 33	0.068 50	0.126 42	0.283 42	0.290 40	0.028 47	0.219 46	0.214 47	0.331 39	0.396 50	0.275 46	0.821 45	0.245 41

MTML	0.549 32	1.000 1	0.807 16	0.588 33	0.327 38	0.647 34	0.004 41	0.815 7	0.180 37	0.418 30	0.364 38	0.182 36	0.445 34	1.000 1	0.442 31	0.688 35	0.571 28	1.000 1	0.396 31
Jean Lahoud, Bernard Ghanem, Marc Pollefeys, Martin R. Oswald: 3D Instance Segmentation via Multi-task Metric Learning. ICCV 2019 [oral]
SegGroup_ins	0.445 41	0.667 35	0.773 25	0.185 47	0.317 39	0.656 33	0.000 44	0.407 41	0.134 41	0.381 36	0.267 41	0.217 35	0.476 32	0.714 37	0.452 30	0.629 39	0.514 36	1.000 1	0.222 43
An Tao, Yueqi Duan, Yi Wei, Jiwen Lu, Jie Zhou: SegGroup: Seg-Level Supervision for 3D Instance and Semantic Segmentation. TIP 2022
3D-BoNet	0.488 36	1.000 1	0.672 41	0.590 32	0.301 40	0.484 47	0.098 28	0.620 26	0.306 32	0.341 39	0.259 42	0.125 40	0.434 37	0.796 36	0.402 34	0.499 46	0.513 37	0.909 39	0.439 28
Bo Yang, Jianan Wang, Ronald Clark, Qingyong Hu, Sen Wang, Andrew Markham, Niki Trigoni: Learning Object Bounding Boxes for 3D Instance Segmentation on Point Clouds. NeurIPS 2019 Spotlight
Region-18class	0.284 46	0.250 51	0.751 32	0.228 45	0.270 41	0.521 44	0.000 44	0.468 39	0.008 50	0.205 45	0.127 46	0.000 51	0.068 50	0.070 50	0.262 43	0.652 37	0.323 43	0.740 46	0.173 45

SALoss-ResNet	0.459 39	1.000 1	0.737 33	0.159 50	0.259 42	0.587 39	0.138 24	0.475 38	0.217 36	0.416 31	0.408 33	0.128 39	0.315 43	0.714 37	0.411 33	0.536 45	0.590 24	0.873 43	0.304 37
Zhidong Liang, Ming Yang, Hao Li, Chunxiang Wang: 3D Instance Embedding Learning With a Structure-Aware Loss Function for Point Cloud Segmentation. IEEE Robotics and Automation Letters (IROS2020)
PanopticFusion-inst	0.478 37	0.667 35	0.712 37	0.595 31	0.259 43	0.550 43	0.000 44	0.613 27	0.175 39	0.250 44	0.434 29	0.437 22	0.411 39	0.857 27	0.485 28	0.591 43	0.267 49	0.944 36	0.359 34
Gaku Narita, Takashi Seno, Tomoya Ishikawa, Yohsuke Kaji: PanopticFusion: Online Volumetric Semantic Mapping at the Level of Stuff and Things. IROS 2019 (to appear)
3D-SIS	0.382 42	1.000 1	0.432 48	0.245 42	0.190 44	0.577 40	0.013 39	0.263 44	0.033 47	0.320 40	0.240 43	0.075 42	0.422 38	0.857 27	0.117 46	0.699 30	0.271 48	0.883 42	0.235 42
Ji Hou, Angela Dai, Matthias Niessner: 3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans. CVPR 2019
UNet-backbone	0.319 44	0.667 35	0.715 36	0.233 43	0.189 45	0.479 48	0.008 40	0.218 45	0.067 46	0.201 46	0.173 45	0.107 41	0.123 48	0.438 44	0.150 44	0.615 41	0.355 41	0.916 38	0.093 51

SemRegionNet-20cls	0.250 47	0.333 48	0.613 42	0.229 44	0.163 46	0.493 45	0.000 44	0.304 43	0.107 43	0.147 48	0.100 47	0.052 45	0.231 44	0.119 48	0.039 48	0.445 48	0.325 42	0.654 47	0.141 47

Hier3D	0.323 43	0.667 35	0.542 45	0.264 41	0.157 47	0.550 42	0.000 44	0.205 47	0.009 48	0.270 43	0.218 44	0.075 42	0.500 30	0.688 39	0.007 52	0.698 31	0.301 45	0.459 49	0.200 44
Tan: HCFS3D: Hierarchical Coupled Feature Selection Network for 3D Semantic and Instance Segmentation.
tmp	0.248 48	0.667 35	0.437 47	0.188 46	0.153 48	0.491 46	0.000 44	0.208 46	0.094 45	0.153 47	0.099 48	0.057 44	0.217 47	0.119 48	0.039 48	0.466 47	0.302 44	0.640 48	0.140 48

ASIS	0.199 50	0.333 48	0.253 51	0.167 49	0.140 49	0.438 49	0.000 44	0.177 48	0.008 49	0.121 49	0.069 49	0.004 50	0.231 45	0.429 45	0.036 50	0.445 49	0.273 47	0.333 51	0.119 50

Sgpn_scannet	0.143 51	0.208 52	0.390 50	0.169 48	0.065 50	0.275 51	0.029 37	0.069 49	0.000 51	0.087 51	0.043 50	0.014 49	0.027 52	0.000 51	0.112 47	0.351 51	0.168 51	0.438 50	0.138 49

MaskRCNN 2d->3d Proj	0.058 52	0.333 48	0.002 52	0.000 52	0.053 51	0.002 52	0.002 43	0.021 52	0.000 51	0.045 52	0.024 52	0.238 34	0.065 51	0.000 51	0.014 51	0.107 52	0.020 52	0.110 52	0.006 52

3D-BEVIS	0.248 48	0.667 35	0.566 43	0.076 51	0.035 52	0.394 50	0.027 38	0.035 51	0.098 44	0.099 50	0.030 51	0.025 48	0.098 49	0.375 46	0.126 45	0.604 42	0.181 50	0.854 44	0.171 46
Cathrin Elich, Francis Engelmann, Jonas Schult, Theodora Kontogianni, Bastian Leibe: 3D-BEVIS: Birds-Eye-View Instance Segmentation.

Method	avg iou	bathtub	bed	bookshelf	cabinet	chair	counter	curtain	desk	door	floor	otherfurniture	picture	refrigerator	shower curtain	sink	sofa	table	toilet	wall	window

Virtual MVFusion (R)	0.745 1	0.861 1	0.839 1	0.881 1	0.672 1	0.512 1	0.422 14	0.898 1	0.723 1	0.714 1	0.954 2	0.454 1	0.509 1	0.773 1	0.895 1	0.756 1	0.820 1	0.653 1	0.935 1	0.891 1	0.728 1
Abhijit Kundu, Xiaoqi Yin, Alireza Fathi, David Ross, Brian Brewington, Thomas Funkhouser, Caroline Pantofaru: Virtual Multi-view Fusion for 3D Semantic Segmentation. ECCV 2020
BPNet_2D	0.670 2	0.822 3	0.795 3	0.836 2	0.659 2	0.481 2	0.451 10	0.769 3	0.656 3	0.567 3	0.931 3	0.395 4	0.390 4	0.700 3	0.534 3	0.689 8	0.770 2	0.574 3	0.865 6	0.831 3	0.675 4
Wenbo Hu, Hengshuang Zhao, Li Jiang, Jiaya Jia and Tien-Tsin Wong: Bidirectional Projection Network for Cross Dimension Scene Understanding. CVPR 2021 (Oral)
CU-Hybrid-2D Net	0.636 3	0.825 2	0.820 2	0.179 19	0.648 3	0.463 3	0.549 2	0.742 6	0.676 2	0.628 2	0.961 1	0.420 2	0.379 5	0.684 6	0.381 14	0.732 2	0.723 3	0.599 2	0.827 12	0.851 2	0.634 6

DMMF_3d	0.605 5	0.651 8	0.744 7	0.782 3	0.637 4	0.387 4	0.536 3	0.732 7	0.590 6	0.540 5	0.856 17	0.359 9	0.306 12	0.596 11	0.539 2	0.627 17	0.706 4	0.497 7	0.785 17	0.757 15	0.476 18

CMX	0.613 4	0.681 7	0.725 8	0.502 11	0.634 5	0.297 15	0.478 8	0.830 2	0.651 4	0.537 6	0.924 4	0.375 5	0.315 11	0.686 5	0.451 11	0.714 4	0.543 18	0.504 5	0.894 4	0.823 4	0.688 3

MIX6D_RVC	0.582 10	0.695 4	0.687 13	0.225 17	0.632 6	0.328 10	0.550 1	0.748 5	0.623 5	0.494 13	0.890 10	0.350 11	0.254 19	0.688 4	0.454 9	0.716 3	0.597 14	0.489 8	0.881 5	0.768 12	0.575 11

FAN_NV_RVC	0.586 8	0.510 17	0.764 4	0.079 22	0.620 7	0.330 8	0.494 6	0.753 4	0.573 8	0.556 4	0.884 12	0.405 3	0.303 13	0.718 2	0.452 10	0.672 11	0.658 5	0.509 4	0.898 3	0.813 6	0.727 2

MCA-Net	0.595 6	0.533 16	0.756 6	0.746 4	0.590 8	0.334 7	0.506 5	0.670 11	0.587 7	0.500 10	0.905 8	0.366 8	0.352 8	0.601 10	0.506 5	0.669 14	0.648 7	0.501 6	0.839 11	0.769 11	0.516 17

RFBNet	0.592 7	0.616 9	0.758 5	0.659 5	0.581 9	0.330 8	0.469 9	0.655 14	0.543 12	0.524 7	0.924 4	0.355 10	0.336 10	0.572 13	0.479 7	0.671 12	0.648 7	0.480 9	0.814 15	0.814 5	0.614 9

segfomer with 6d	0.542 15	0.594 12	0.687 13	0.146 20	0.579 10	0.308 13	0.515 4	0.703 10	0.472 17	0.498 11	0.868 14	0.369 7	0.282 14	0.589 12	0.390 13	0.701 7	0.556 17	0.416 18	0.860 9	0.759 14	0.539 15

SSMA	0.577 11	0.695 4	0.716 11	0.439 13	0.563 11	0.314 11	0.444 12	0.719 8	0.551 10	0.503 9	0.887 11	0.346 12	0.348 9	0.603 9	0.353 16	0.709 5	0.600 12	0.457 12	0.901 2	0.786 8	0.599 10
Abhinav Valada, Rohit Mohan, Wolfram Burgard: Self-Supervised Model Adaptation for Multimodal Semantic Segmentation. International Journal of Computer Vision, 2019
UDSSEG_RVC	0.545 14	0.610 11	0.661 18	0.588 8	0.556 12	0.268 18	0.482 7	0.642 16	0.572 9	0.475 14	0.836 19	0.312 14	0.367 6	0.630 7	0.189 19	0.639 16	0.495 20	0.452 13	0.826 13	0.756 16	0.541 13

UNIV_CNP_RVC_UE	0.566 12	0.569 15	0.686 15	0.435 14	0.524 13	0.294 16	0.421 15	0.712 9	0.543 12	0.463 15	0.872 13	0.320 13	0.363 7	0.611 8	0.477 8	0.686 9	0.627 9	0.443 15	0.862 7	0.775 10	0.639 5

SN_RN152pyrx8_RVC	0.546 13	0.572 13	0.663 17	0.638 7	0.518 14	0.298 14	0.366 20	0.633 17	0.510 15	0.446 17	0.864 15	0.296 16	0.267 16	0.542 15	0.346 17	0.704 6	0.575 16	0.431 16	0.853 10	0.766 13	0.630 7

FuseNet	0.535 16	0.570 14	0.681 16	0.182 18	0.512 15	0.290 17	0.431 13	0.659 13	0.504 16	0.495 12	0.903 9	0.308 15	0.428 3	0.523 17	0.365 15	0.676 10	0.621 11	0.470 11	0.762 18	0.779 9	0.541 13
Caner Hazirbas, Lingni Ma, Csaba Domokos, Daniel Cremers: FuseNet: Incorporating Depth into Semantic Segmentation via Fusion-based CNN Architecture. ACCV 2016
DCRedNet	0.583 9	0.682 6	0.723 9	0.542 10	0.510 16	0.310 12	0.451 10	0.668 12	0.549 11	0.520 8	0.920 6	0.375 5	0.446 2	0.528 16	0.417 12	0.670 13	0.577 15	0.478 10	0.862 7	0.806 7	0.628 8

ILC-PSPNet	0.475 20	0.490 19	0.581 20	0.289 16	0.507 17	0.067 22	0.379 18	0.610 19	0.417 20	0.435 18	0.822 21	0.278 17	0.267 16	0.503 18	0.228 18	0.616 19	0.533 19	0.375 19	0.820 14	0.729 17	0.560 12

3DMV (2d proj)	0.498 18	0.481 20	0.612 19	0.579 9	0.456 18	0.343 5	0.384 17	0.623 18	0.525 14	0.381 19	0.845 18	0.254 18	0.264 18	0.557 14	0.182 20	0.581 20	0.598 13	0.429 17	0.760 19	0.661 21	0.446 20
Angela Dai, Matthias Niessner: 3DMV: Joint 3D-Multi-View Prediction for 3D Semantic Scene Segmentation. ECCV'18
MSeg1080_RVC	0.485 19	0.505 18	0.709 12	0.092 21	0.427 19	0.241 19	0.411 16	0.654 15	0.385 22	0.457 16	0.861 16	0.053 22	0.279 15	0.503 18	0.481 6	0.645 15	0.626 10	0.365 20	0.748 20	0.725 18	0.529 16
John Lambert, Zhuang Liu, Ozan Sener, James Hays, Vladlen Koltun: MSeg: A Composite Dataset for Multi-domain Semantic Segmentation. CVPR 2020
Enet (reimpl)	0.376 21	0.264 22	0.452 22	0.452 12	0.365 20	0.181 20	0.143 22	0.456 21	0.409 21	0.346 21	0.769 22	0.164 20	0.218 20	0.359 21	0.123 22	0.403 22	0.381 22	0.313 22	0.571 21	0.685 20	0.472 19
Re-implementation of Adam Paszke, Abhishek Chaurasia, Sangpil Kim, Eugenio Culurciello: ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation.
ScanNet (2d proj)	0.330 22	0.293 21	0.521 21	0.657 6	0.361 21	0.161 21	0.250 21	0.004 22	0.440 19	0.183 22	0.836 19	0.125 21	0.060 22	0.319 22	0.132 21	0.417 21	0.412 21	0.344 21	0.541 22	0.427 22	0.109 22
Angela Dai, Angel X. Chang, Manolis Savva, Maciej Halber, Thomas Funkhouser, Matthias Nießner: ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes. CVPR'17
AdapNet++	0.503 17	0.613 10	0.722 10	0.418 15	0.358 22	0.337 6	0.370 19	0.479 20	0.443 18	0.368 20	0.907 7	0.207 19	0.213 21	0.464 20	0.525 4	0.618 18	0.657 6	0.450 14	0.788 16	0.721 19	0.408 21
Abhinav Valada, Rohit Mohan, Wolfram Burgard: Self-Supervised Model Adaptation for Multimodal Semantic Segmentation. International Journal of Computer Vision, 2019
DMMF	0.003 23	0.000 23	0.005 23	0.000 23	0.000 23	0.037 23	0.001 23	0.000 23	0.001 23	0.005 23	0.003 23	0.000 23	0.000 23	0.000 23	0.000 23	0.000 23	0.002 23	0.001 23	0.000 23	0.006 23	0.000 23

Presenting the ScanNet200 Benchmark

ScanNet200 Benchmark

ScanNet200 3D Semantic Label Benchmark

ScanNet200 3D Semantic Instance Benchmark

ScanNet Benchmark

3D Semantic Label Benchmark

3D Semantic Instance Benchmark

2D Semantic Label Benchmark

2D Semantic Instance Benchmark

Scene Type Classification Benchmark

Method	avg ap	bathtub	bed	bookshelf	cabinet	chair	counter	curtain	desk	door	otherfurniture	picture	refrigerator	shower curtain	sink	sofa	table	toilet	window

UniDet_RVC	0.205 1	0.381 1	0.323 1	0.037 1	0.226 1	0.177 1	0.063 1	0.277 1	0.120 1	0.067 1	0.131 1	0.074 2	0.317 1	0.080 1	0.235 1	0.289 1	0.141 1	0.678 1	0.080 1

MaskRCNN_ScanNet	0.119 2	0.129 2	0.212 2	0.002 2	0.112 2	0.148 2	0.014 2	0.205 2	0.044 2	0.066 2	0.078 2	0.095 1	0.142 2	0.030 2	0.128 2	0.139 2	0.080 2	0.459 2	0.057 2
Re-implementation of Kaiming He, Georgia Gkioxari, Piotr Dollár, Ross Girshick: Mask R-CNN. ICCV'17

Method	avg recall	apartment	bathroom	bedroom / hotel	bookstore / library	conference room	copy/mail room	hallway	kitchen	laundry room	living room / lounge	misc	office	storage / basement / garage

SE-ResNeXt-SSMA	0.498 3	0.000 4	0.812 3	0.941 1	0.500 2	0.500 3	0.500 2	0.500 2	0.429 4	0.500 2	0.667 2	0.500 1	0.625 3	0.000 2
Abhinav Valada, Rohit Mohan, Wolfram Burgard: Self-Supervised Model Adaptation for Multimodal Semantic Segmentation. arXiv
multi-task	0.700 1	0.500 1	1.000 1	0.882 2	0.500 2	1.000 1	1.000 1	0.500 2	1.000 1	1.000 1	0.778 1	0.000 2	0.938 1	0.000 2
Shengyu Huang, Mikhail Usvyatsov, Konrad Schindler: Indoor Scene Recognition in 3D. IROS 2020
3DASPP-SCE	0.691 2	0.500 1	0.938 2	0.824 3	1.000 1	1.000 1	0.500 2	1.000 1	0.857 2	0.500 2	0.556 3	0.000 2	0.812 2	0.500 1

resnet50_scannet	0.353 4	0.250 3	0.812 3	0.529 4	0.500 2	0.500 3	0.000 4	0.500 2	0.571 3	0.000 4	0.556 3	0.000 2	0.375 4	0.000 2