3D Semantic Instance Results - ScanNet Benchmark

The 3D semantic instance prediction task involves detecting and segmenting the object in an 3D scan mesh.

Evaluation and metrics

Our evaluation ranks all methods according to the average precision for each class. We report the mean average precision AP at overlap 0.25 (AP 25%), overlap 0.5 (AP 50%), and over overlaps in the range [0.5:0.95:0.05] (AP). Note that multiple predictions of the same ground truth instance are penalized as false positives.

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

Method	avg ap 25%	bathtub	bed	bookshelf	cabinet	chair	counter	curtain	desk	door	otherfurniture	picture	refrigerator	shower curtain	sink	sofa	table	toilet	window

PointRel	0.901 1	1.000 1	0.978 23	0.928 3	0.879 1	0.962 5	0.882 5	0.749 38	0.947 3	0.912 2	0.802 3	0.753 19	0.820 2	1.000 1	0.984 4	0.919 5	0.894 4	1.000 1	0.815 15
: Relation3D (PointRel): Enhancing Relation Modeling for Point Cloud Instance Segmentation.
Competitor-MAFT	0.896 2	1.000 1	1.000 1	0.872 16	0.847 11	0.967 3	0.955 1	0.778 33	0.901 15	0.919 1	0.784 5	0.812 1	0.770 13	1.000 1	0.949 9	0.865 35	0.868 18	1.000 1	0.840 5

OneFormer3D	0.896 2	1.000 1	1.000 1	0.913 6	0.858 6	0.951 10	0.786 16	0.837 19	0.916 12	0.908 4	0.778 8	0.803 6	0.750 15	1.000 1	0.976 6	0.926 4	0.882 8	0.995 48	0.849 2
Maxim Kolodiazhnyi, Anna Vorontsova, Anton Konushin, Danila Rukhovich: OneFormer3D: One Transformer for Unified Point Cloud Segmentation.
MG-Former	0.887 4	1.000 1	0.991 14	0.837 26	0.801 25	0.935 19	0.887 4	0.857 11	0.946 4	0.891 10	0.748 18	0.805 5	0.739 17	1.000 1	0.993 2	0.809 59	0.876 15	1.000 1	0.842 4

DCD	0.885 5	1.000 1	0.933 41	0.856 22	0.832 15	0.959 7	0.930 2	0.858 10	0.802 38	0.859 18	0.767 9	0.796 10	0.709 21	1.000 1	0.971 7	0.871 29	0.904 2	1.000 1	0.874 1

UniPerception	0.884 6	1.000 1	0.979 20	0.872 16	0.869 3	0.892 28	0.806 13	0.890 6	0.835 29	0.892 9	0.755 14	0.811 2	0.779 10	0.955 49	0.951 8	0.876 23	0.914 1	0.997 40	0.840 6

KmaxOneFormerNet	0.883 7	1.000 1	1.000 1	0.798 41	0.848 10	0.971 1	0.853 7	0.903 3	0.827 32	0.910 3	0.748 17	0.809 4	0.724 19	1.000 1	0.980 5	0.855 41	0.844 24	1.000 1	0.832 7

InsSSM	0.883 7	1.000 1	0.996 6	0.800 40	0.865 4	0.960 6	0.808 12	0.852 16	0.940 6	0.899 8	0.785 4	0.810 3	0.700 23	1.000 1	0.912 20	0.851 44	0.895 3	0.997 40	0.827 9
Lei Yao, Yi Wang, Moyun Liu, Lap-Pui Chau: SGIFormer: Semantic-guided and Geometric-enhanced Interleaving Transformer for 3D Instance Segmentation. TCSVT, 2024
Competitor-SPFormer	0.881 9	1.000 1	1.000 1	0.845 24	0.854 7	0.962 4	0.714 23	0.857 12	0.904 14	0.902 6	0.782 7	0.789 13	0.662 29	1.000 1	0.988 3	0.874 26	0.886 7	0.997 40	0.847 3

TST3D	0.879 10	1.000 1	0.994 9	0.921 5	0.807 24	0.939 16	0.771 17	0.887 7	0.923 10	0.862 17	0.722 23	0.768 16	0.756 14	1.000 1	0.910 31	0.904 7	0.836 27	0.999 39	0.824 11
Duc Tran Dang Trung, Byeongkeun Kang, Yeejin Lee: MSTA3D: Multi-scale Twin-attention for 3D Instance Segmentation. ACM Multimedia 2024
SIM3D	0.878 11	1.000 1	0.972 25	0.863 19	0.817 22	0.952 9	0.821 10	0.783 30	0.890 18	0.902 7	0.735 21	0.797 8	0.799 9	1.000 1	0.931 17	0.893 13	0.853 22	1.000 1	0.792 18

EV3D	0.877 12	1.000 1	0.996 8	0.873 14	0.854 8	0.950 11	0.691 27	0.783 31	0.926 7	0.889 13	0.754 15	0.794 12	0.820 2	1.000 1	0.912 20	0.900 9	0.860 20	1.000 1	0.779 21

Spherical Mask(CtoF)	0.875 13	1.000 1	0.991 15	0.873 14	0.850 9	0.946 13	0.691 27	0.752 37	0.926 7	0.889 12	0.759 12	0.794 11	0.820 2	1.000 1	0.912 20	0.900 9	0.878 12	1.000 1	0.769 23

TD3D	0.875 13	1.000 1	0.976 24	0.877 12	0.783 31	0.970 2	0.889 3	0.828 20	0.945 5	0.803 24	0.713 25	0.720 26	0.709 20	1.000 1	0.936 15	0.934 3	0.873 16	1.000 1	0.791 19
Maksim Kolodiazhnyi, Anna Vorontsova, Anton Konushin, Danila Rukhovich: Top-Down Beats Bottom-Up in 3D Instance Segmentation. WACV 2024
SoftGroup++	0.874 15	1.000 1	0.972 26	0.947 1	0.839 14	0.898 27	0.556 42	0.913 2	0.881 21	0.756 26	0.828 2	0.748 21	0.821 1	1.000 1	0.937 14	0.937 1	0.887 6	1.000 1	0.821 12

Queryformer	0.874 15	1.000 1	0.978 22	0.809 38	0.876 2	0.936 18	0.702 24	0.716 43	0.920 11	0.875 16	0.766 10	0.772 15	0.818 6	1.000 1	0.995 1	0.916 6	0.892 5	1.000 1	0.767 24

Mask3D	0.870 17	1.000 1	0.985 17	0.782 48	0.818 21	0.938 17	0.760 18	0.749 38	0.923 9	0.877 15	0.760 11	0.785 14	0.820 2	1.000 1	0.912 20	0.864 37	0.878 12	0.983 54	0.825 10
Jonas Schult, Francis Engelmann, Alexander Hermans, Or Litany, Siyu Tang, Bastian Leibe: Mask3D for 3D Semantic Instance Segmentation. ICRA 2023
ExtMask3D	0.867 18	1.000 1	1.000 1	0.756 55	0.816 23	0.940 15	0.795 14	0.760 36	0.862 23	0.888 14	0.739 19	0.763 17	0.774 11	1.000 1	0.929 18	0.878 22	0.879 10	1.000 1	0.819 14

SoftGroup	0.865 19	1.000 1	0.969 27	0.860 20	0.860 5	0.913 23	0.558 39	0.899 4	0.911 13	0.760 25	0.828 1	0.736 23	0.802 8	0.981 46	0.919 19	0.875 24	0.877 14	1.000 1	0.820 13
Thang Vu, Kookhoi Kim, Tung M. Luu, Xuan Thanh Nguyen, Chang D. Yoo: SoftGroup for 3D Instance Segmentaiton on Point Clouds. CVPR 2022 [Oral]
MAFT	0.860 20	1.000 1	0.990 16	0.810 37	0.829 16	0.949 12	0.809 11	0.688 49	0.836 28	0.904 5	0.751 16	0.796 9	0.741 16	1.000 1	0.864 41	0.848 46	0.837 25	1.000 1	0.828 8

SPFormer	0.851 21	1.000 1	0.994 10	0.806 39	0.774 33	0.942 14	0.637 31	0.849 17	0.859 25	0.889 11	0.720 24	0.730 24	0.665 28	1.000 1	0.911 28	0.868 34	0.873 17	1.000 1	0.796 17
Sun Jiahao, Qing Chunmei, Tan Junpeng, Xu Xiangmin: Superpoint Transformer for 3D Scene Instance Segmentation. AAAI 2023 [Oral]
IPCA-Inst	0.851 21	1.000 1	0.968 28	0.884 11	0.842 13	0.862 41	0.693 26	0.812 25	0.888 20	0.677 38	0.783 6	0.698 27	0.807 7	1.000 1	0.911 28	0.865 36	0.865 19	1.000 1	0.757 27

ODIN - Ins	0.847 23	1.000 1	0.951 34	0.834 31	0.828 17	0.875 33	0.871 6	0.767 34	0.821 34	0.816 21	0.690 32	0.800 7	0.771 12	1.000 1	0.912 20	0.891 14	0.821 28	0.886 70	0.713 34
Ayush Jain, Pushkal Katara, Nikolaos Gkanatsios, Adam W. Harley, Gabriel Sarch, Kriti Aggarwal, Vishrav Chaudhary, Katerina Fragkiadaki: ODIN: A Single Model for 2D and 3D Segmentation. CVPR 2024
Mask3D_evaluation	0.843 24	1.000 1	0.955 33	0.847 23	0.795 27	0.932 20	0.750 20	0.780 32	0.891 17	0.818 20	0.737 20	0.633 36	0.703 22	1.000 1	0.902 33	0.870 30	0.820 29	0.941 62	0.805 16

ISBNet	0.835 25	1.000 1	0.950 35	0.731 57	0.819 19	0.918 21	0.790 15	0.740 40	0.851 27	0.831 19	0.661 34	0.742 22	0.650 32	1.000 1	0.937 13	0.814 58	0.836 26	1.000 1	0.765 25
Tuan Duc Ngo, Binh-Son Hua, Khoi Nguyen: ISBNet: a 3D Point Cloud Instance Segmentation Network with Instance-aware Sampling and Box-aware Dynamic Convolution. CVPR 2023
SphereSeg	0.835 25	1.000 1	0.963 31	0.891 9	0.794 28	0.954 8	0.822 9	0.710 44	0.961 2	0.721 30	0.693 31	0.530 49	0.653 31	1.000 1	0.867 40	0.857 40	0.859 21	0.991 51	0.771 22

GraphCut	0.832 27	1.000 1	0.922 50	0.724 59	0.798 26	0.902 26	0.701 25	0.856 14	0.859 24	0.715 31	0.706 26	0.748 20	0.640 43	1.000 1	0.934 16	0.862 38	0.880 9	1.000 1	0.729 30

TopoSeg	0.832 27	1.000 1	0.981 19	0.933 2	0.819 20	0.826 50	0.524 48	0.841 18	0.811 35	0.681 37	0.759 13	0.687 28	0.727 18	0.981 46	0.911 28	0.883 18	0.853 23	1.000 1	0.756 28

PBNet	0.825 29	1.000 1	0.963 30	0.837 28	0.843 12	0.865 36	0.822 8	0.647 52	0.878 22	0.733 28	0.639 41	0.683 29	0.650 32	1.000 1	0.853 42	0.870 31	0.820 30	1.000 1	0.744 29
Weiguang Zhao, Yuyao Yan, Chaolong Yang, Jianan Ye, Xi Yang, Kaizhu Huang: Divide and Conquer: 3D Instance Segmentation With Point-Wise Binarization. ICCV 2023
SSEC	0.820 30	1.000 1	0.983 18	0.924 4	0.826 18	0.817 53	0.415 57	0.899 5	0.793 40	0.673 39	0.731 22	0.636 34	0.653 30	1.000 1	0.939 12	0.804 61	0.878 11	1.000 1	0.780 20

DKNet	0.815 31	1.000 1	0.930 42	0.844 25	0.765 37	0.915 22	0.534 46	0.805 27	0.805 37	0.807 23	0.654 35	0.763 18	0.650 32	1.000 1	0.794 54	0.881 19	0.766 34	1.000 1	0.758 26
Yizheng Wu, Min Shi, Shuaiyuan Du, Hao Lu, Zhiguo Cao, Weicai Zhong: 3D Instances as 1D Kernels. ECCV 2022
RPGN	0.806 32	1.000 1	0.992 12	0.789 43	0.723 50	0.891 29	0.650 30	0.810 26	0.832 30	0.665 41	0.699 29	0.658 30	0.700 23	1.000 1	0.881 35	0.832 50	0.774 32	0.997 40	0.613 51
Shichao Dong, Guosheng Lin, Tzu-Yi Hung: Learning Regional Purity for Instance Segmentation on 3D Point Clouds. ECCV 2022
HAIS	0.803 33	1.000 1	0.994 10	0.820 33	0.759 38	0.855 42	0.554 43	0.882 8	0.827 33	0.615 47	0.676 33	0.638 33	0.646 41	1.000 1	0.912 20	0.797 64	0.767 33	0.994 49	0.726 31
Shaoyu Chen, Jiemin Fang, Qian Zhang, Wenyu Liu, Xinggang Wang: Hierarchical Aggregation for 3D Instance Segmentation. ICCV 2021
Box2Mask	0.803 33	1.000 1	0.962 32	0.874 13	0.707 54	0.887 32	0.686 29	0.598 57	0.961 1	0.715 32	0.694 30	0.469 54	0.700 23	1.000 1	0.912 20	0.902 8	0.753 39	0.997 40	0.637 45
Julian Chibane, Francis Engelmann, Tuan Anh Tran, Gerard Pons-Moll: Box2Mask: Weakly Supervised 3D Semantic Instance Segmentation Using Bounding Boxes. ECCV 2022
Mask-Group	0.792 35	1.000 1	0.968 29	0.812 34	0.766 36	0.864 37	0.460 51	0.815 24	0.888 19	0.598 51	0.651 38	0.639 32	0.600 49	0.918 52	0.941 10	0.896 12	0.721 46	1.000 1	0.723 32
Min Zhong, Xinghao Chen, Xiaokang Chen, Gang Zeng, Yunhe Wang: MaskGroup: Hierarchical Point Grouping and Masking for 3D Instance Segmentation. ICME 2022
CSC-Pretrained	0.791 36	1.000 1	0.996 6	0.829 32	0.767 35	0.889 31	0.600 34	0.819 23	0.770 45	0.594 52	0.620 45	0.541 46	0.700 23	1.000 1	0.941 10	0.889 16	0.763 35	1.000 1	0.526 61

SSTNet	0.789 37	1.000 1	0.840 64	0.888 10	0.717 51	0.835 46	0.717 22	0.684 50	0.627 60	0.724 29	0.652 37	0.727 25	0.600 49	1.000 1	0.912 20	0.822 53	0.757 38	1.000 1	0.691 39
Zhihao Liang, Zhihao Li, Songcen Xu, Mingkui Tan, Kui Jia: Instance Segmentation in 3D Scenes using Semantic Superpoint Tree Networks. ICCV2021
GICN	0.788 38	1.000 1	0.978 21	0.867 18	0.781 32	0.833 47	0.527 47	0.824 21	0.806 36	0.549 60	0.596 48	0.551 42	0.700 23	1.000 1	0.853 42	0.935 2	0.733 43	1.000 1	0.651 42

DENet	0.786 39	1.000 1	0.929 43	0.736 56	0.750 44	0.720 66	0.755 19	0.934 1	0.794 39	0.590 53	0.561 54	0.537 47	0.650 32	1.000 1	0.882 34	0.804 62	0.789 31	1.000 1	0.719 33

DANCENET	0.786 39	1.000 1	0.936 38	0.783 46	0.737 47	0.852 44	0.742 21	0.647 52	0.765 47	0.811 22	0.624 44	0.579 39	0.632 46	1.000 1	0.909 32	0.898 11	0.696 51	0.944 58	0.601 54

DualGroup	0.782 41	1.000 1	0.927 44	0.811 35	0.772 34	0.853 43	0.631 33	0.805 27	0.773 42	0.613 48	0.611 46	0.610 37	0.650 32	0.835 63	0.881 35	0.879 21	0.750 41	1.000 1	0.675 40

PointGroup	0.778 42	1.000 1	0.900 54	0.798 42	0.715 52	0.863 38	0.493 49	0.706 45	0.895 16	0.569 58	0.701 27	0.576 40	0.639 44	1.000 1	0.880 37	0.851 43	0.719 47	0.997 40	0.709 36
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]
PE	0.776 43	1.000 1	0.900 55	0.860 20	0.728 49	0.869 34	0.400 58	0.857 13	0.774 41	0.568 59	0.701 28	0.602 38	0.646 41	0.933 51	0.843 45	0.890 15	0.691 55	0.997 40	0.709 35
Biao Zhang, Peter Wonka: Point Cloud Instance Segmentation using Probabilistic Embeddings. CVPR 2021
AOIA	0.767 44	1.000 1	0.937 37	0.810 36	0.740 46	0.906 24	0.550 44	0.800 29	0.706 52	0.577 57	0.624 43	0.544 45	0.596 54	0.857 55	0.879 39	0.880 20	0.750 40	0.992 50	0.658 41

DD-UNet+Group	0.764 45	1.000 1	0.897 57	0.837 27	0.753 41	0.830 49	0.459 53	0.824 21	0.699 54	0.629 45	0.653 36	0.438 57	0.650 32	1.000 1	0.880 37	0.858 39	0.690 56	1.000 1	0.650 43
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
INS-Conv-instance	0.762 46	1.000 1	0.923 47	0.765 51	0.785 30	0.905 25	0.600 34	0.655 51	0.646 59	0.683 36	0.647 39	0.530 48	0.650 32	1.000 1	0.824 47	0.830 51	0.693 54	0.944 58	0.644 44

Dyco3D	0.761 47	1.000 1	0.935 39	0.893 8	0.752 43	0.863 39	0.600 34	0.588 58	0.742 49	0.641 43	0.633 42	0.546 44	0.550 56	0.857 55	0.789 56	0.853 42	0.762 36	0.987 52	0.699 37
Tong He; Chunhua Shen; Anton van den Hengel: DyCo3D: Robust Instance Segmentation of 3D Point Clouds through Dynamic Convolution. CVPR2021
OccuSeg+instance	0.742 48	1.000 1	0.923 47	0.785 44	0.745 45	0.867 35	0.557 40	0.578 61	0.729 50	0.670 40	0.644 40	0.488 52	0.577 55	1.000 1	0.794 54	0.830 51	0.620 64	1.000 1	0.550 57
Lei Han, Tian Zheng, Lan Xu, Lu Fang: OccuSeg: Occupancy-aware 3D Instance Segmentation. CVPR2020
RWSeg	0.739 49	1.000 1	0.899 56	0.759 53	0.753 42	0.823 51	0.282 63	0.691 48	0.658 57	0.582 56	0.594 49	0.547 43	0.628 47	1.000 1	0.795 53	0.868 33	0.728 45	1.000 1	0.692 38

3D-MPA	0.737 50	1.000 1	0.933 40	0.785 44	0.794 29	0.831 48	0.279 65	0.588 58	0.695 55	0.616 46	0.559 55	0.556 41	0.650 32	1.000 1	0.809 51	0.875 25	0.696 52	1.000 1	0.608 53
Francis Engelmann, Martin Bokeloh, Alireza Fathi, Bastian Leibe, Matthias Nießner: 3D-MPA: Multi Proposal Aggregation for 3D Semantic Instance Segmentation. CVPR 2020
MTML	0.731 51	1.000 1	0.992 12	0.779 50	0.609 63	0.746 61	0.308 62	0.867 9	0.601 63	0.607 49	0.539 58	0.519 50	0.550 56	1.000 1	0.824 47	0.869 32	0.729 44	1.000 1	0.616 49
Jean Lahoud, Bernard Ghanem, Marc Pollefeys, Martin R. Oswald: 3D Instance Segmentation via Multi-task Metric Learning. ICCV 2019 [oral]
OSIS	0.725 52	1.000 1	0.885 60	0.653 65	0.657 60	0.801 54	0.576 38	0.695 47	0.828 31	0.698 34	0.534 59	0.457 56	0.500 63	0.857 55	0.831 46	0.841 48	0.627 62	1.000 1	0.619 48

SSEN	0.724 53	1.000 1	0.926 45	0.781 49	0.661 58	0.845 45	0.596 37	0.529 64	0.764 48	0.653 42	0.489 65	0.461 55	0.500 63	0.859 54	0.765 57	0.872 28	0.761 37	1.000 1	0.577 55
Dongsu Zhang, Junha Chun, Sang Kyun Cha, Young Min Kim: Spatial Semantic Embedding Network: Fast 3D Instance Segmentation with Deep Metric Learning. Arxiv
NeuralBF	0.718 54	1.000 1	0.945 36	0.901 7	0.754 40	0.817 52	0.460 51	0.700 46	0.772 43	0.688 35	0.568 53	0.000 76	0.500 63	0.981 46	0.606 67	0.872 27	0.740 42	1.000 1	0.614 50
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
Sparse R-CNN	0.714 55	1.000 1	0.926 46	0.694 60	0.699 56	0.890 30	0.636 32	0.516 65	0.693 56	0.743 27	0.588 50	0.369 61	0.601 48	0.594 69	0.800 52	0.886 17	0.676 57	0.986 53	0.546 58

SALoss-ResNet	0.695 56	1.000 1	0.855 62	0.579 70	0.589 65	0.735 64	0.484 50	0.588 58	0.856 26	0.634 44	0.571 52	0.298 62	0.500 63	1.000 1	0.824 47	0.818 54	0.702 50	0.935 65	0.545 59
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	0.693 57	1.000 1	0.852 63	0.655 64	0.616 62	0.788 56	0.334 60	0.763 35	0.771 44	0.457 70	0.555 56	0.652 31	0.518 60	0.857 55	0.765 57	0.732 70	0.631 60	0.944 58	0.577 56
Gaku Narita, Takashi Seno, Tomoya Ishikawa, Yohsuke Kaji: PanopticFusion: Online Volumetric Semantic Mapping at the Level of Stuff and Things. IROS 2019 (to appear)
Occipital-SCS	0.688 58	1.000 1	0.913 51	0.730 58	0.737 48	0.743 63	0.442 54	0.855 15	0.655 58	0.546 61	0.546 57	0.263 64	0.508 62	0.889 53	0.568 68	0.771 67	0.705 49	0.889 68	0.625 47

3D-BoNet	0.687 59	1.000 1	0.887 59	0.836 29	0.587 66	0.643 73	0.550 44	0.620 54	0.724 51	0.522 65	0.501 63	0.243 65	0.512 61	1.000 1	0.751 59	0.807 60	0.661 59	0.909 67	0.612 52
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
ClickSeg_Instance	0.685 60	1.000 1	0.818 66	0.600 68	0.715 53	0.795 55	0.557 40	0.533 63	0.591 65	0.601 50	0.519 61	0.429 59	0.638 45	0.938 50	0.706 62	0.817 56	0.624 63	0.944 58	0.502 63

PCJC	0.684 61	1.000 1	0.895 58	0.757 54	0.659 59	0.862 40	0.189 72	0.739 41	0.606 62	0.712 33	0.581 51	0.515 51	0.650 32	0.857 55	0.357 73	0.785 65	0.631 61	0.889 68	0.635 46

SPG_WSIS	0.678 62	1.000 1	0.880 61	0.836 29	0.701 55	0.727 65	0.273 67	0.607 56	0.706 53	0.541 63	0.515 62	0.174 68	0.600 49	0.857 55	0.716 61	0.846 47	0.711 48	1.000 1	0.506 62

One_Thing_One_Click	0.675 63	1.000 1	0.823 65	0.782 47	0.621 61	0.766 58	0.211 69	0.736 42	0.560 67	0.586 54	0.522 60	0.636 35	0.453 67	0.641 67	0.853 42	0.850 45	0.694 53	0.997 40	0.411 68
Zhengzhe Liu, Xiaojuan Qi, Chi-Wing Fu: One Thing One Click: A Self-Training Approach for Weakly Supervised 3D Semantic Segmentation. CVPR 2021
SegGroup_ins	0.637 64	1.000 1	0.923 49	0.593 69	0.561 67	0.746 62	0.143 74	0.504 66	0.766 46	0.485 68	0.442 66	0.372 60	0.530 59	0.714 64	0.815 50	0.775 66	0.673 58	1.000 1	0.431 67
An Tao, Yueqi Duan, Yi Wei, Jiwen Lu, Jie Zhou: SegGroup: Seg-Level Supervision for 3D Instance and Semantic Segmentation. TIP 2022
MASC	0.615 65	0.711 72	0.802 67	0.540 71	0.757 39	0.777 57	0.029 75	0.577 62	0.588 66	0.521 66	0.600 47	0.436 58	0.534 58	0.697 65	0.616 66	0.838 49	0.526 66	0.980 55	0.534 60
Chen Liu, Yasutaka Furukawa: MASC: Multi-scale Affinity with Sparse Convolution for 3D Instance Segmentation.
UNet-backbone	0.605 66	1.000 1	0.909 52	0.764 52	0.603 64	0.704 67	0.415 56	0.301 71	0.548 68	0.461 69	0.394 67	0.267 63	0.386 69	0.857 55	0.649 65	0.817 55	0.504 68	0.959 56	0.356 71

3D-SIS	0.558 67	1.000 1	0.773 68	0.614 67	0.503 70	0.691 69	0.200 70	0.412 67	0.498 71	0.546 62	0.311 72	0.103 72	0.600 49	0.857 55	0.382 70	0.799 63	0.445 74	0.938 64	0.371 69
Ji Hou, Angela Dai, Matthias Niessner: 3D-SIS: 3D Semantic Instance Segmentation of RGB-D Scans. CVPR 2019
R-PointNet	0.544 68	0.500 75	0.655 74	0.661 63	0.663 57	0.765 59	0.432 55	0.214 74	0.612 61	0.584 55	0.499 64	0.204 67	0.286 73	0.429 72	0.655 64	0.650 75	0.539 65	0.950 57	0.499 64

Hier3D	0.540 69	1.000 1	0.727 69	0.626 66	0.467 73	0.693 68	0.200 70	0.412 67	0.480 72	0.528 64	0.318 71	0.077 75	0.600 49	0.688 66	0.382 70	0.768 68	0.472 70	0.941 62	0.350 72
Tan: HCFS3D: Hierarchical Coupled Feature Selection Network for 3D Semantic and Instance Segmentation.
Region-18class	0.497 70	0.250 77	0.902 53	0.689 61	0.540 68	0.747 60	0.276 66	0.610 55	0.268 76	0.489 67	0.348 68	0.000 76	0.243 76	0.220 75	0.663 63	0.814 57	0.459 72	0.928 66	0.496 65

Sem_Recon_ins	0.484 71	0.764 71	0.608 76	0.470 73	0.521 69	0.637 74	0.311 61	0.218 73	0.348 75	0.365 74	0.223 73	0.222 66	0.258 74	0.629 68	0.734 60	0.596 76	0.509 67	0.858 72	0.444 66

tmp	0.474 72	1.000 1	0.727 69	0.433 75	0.481 72	0.673 71	0.022 77	0.380 69	0.517 70	0.436 72	0.338 70	0.128 70	0.343 71	0.429 72	0.291 75	0.728 71	0.473 69	0.833 73	0.300 74

SemRegionNet-20cls	0.470 73	1.000 1	0.727 69	0.447 74	0.481 71	0.678 70	0.024 76	0.380 69	0.518 69	0.440 71	0.339 69	0.128 70	0.350 70	0.429 72	0.212 76	0.711 72	0.465 71	0.833 73	0.290 75

ASIS	0.422 74	0.333 76	0.707 72	0.676 62	0.401 74	0.650 72	0.350 59	0.177 75	0.594 64	0.376 73	0.202 74	0.077 74	0.404 68	0.571 70	0.197 77	0.674 74	0.447 73	0.500 76	0.260 76

3D-BEVIS	0.401 75	0.667 73	0.687 73	0.419 76	0.137 77	0.587 75	0.188 73	0.235 72	0.359 74	0.211 76	0.093 77	0.080 73	0.311 72	0.571 70	0.382 70	0.754 69	0.300 76	0.874 71	0.357 70
Cathrin Elich, Francis Engelmann, Jonas Schult, Theodora Kontogianni, Bastian Leibe: 3D-BEVIS: Birds-Eye-View Instance Segmentation.
Sgpn_scannet	0.390 76	0.556 74	0.636 75	0.493 72	0.353 75	0.539 76	0.271 68	0.160 76	0.450 73	0.359 75	0.178 75	0.146 69	0.250 75	0.143 76	0.347 74	0.698 73	0.436 75	0.667 75	0.331 73

MaskRCNN 2d->3d Proj	0.261 77	0.903 70	0.081 77	0.008 77	0.233 76	0.175 77	0.280 64	0.106 77	0.150 77	0.203 77	0.175 76	0.480 53	0.218 77	0.143 76	0.542 69	0.404 77	0.153 77	0.393 77	0.049 77

3D Semantic Instance Benchmark