3D Instance Segmentation on SUNCG synthetic scans benchmark leaderboard · SOTA2 Research

3D Instance Segmentation on SUNCG synthetic scans

30.6mAP@0.253D-SIS

Evaluation Results

Method	Links
3D-SIS2018.12		30.6	20.5	69.4	56.2	64.5	43.8	17.8	0	30	32.3	33.5	21	0	34.2	0	10	20	56.7	56.2	17.6	56.2	10	35.5	17.8
3D-SIS2018.12		29.4	26.5	78.4	48.2	59.5	42.8	26.1	0	30	22.7	39.4	17.3	0	36.2	0	10	10	37	50.8	16.8	59.3	10	36.4	17.8
3D-SIS2018.12		29.1	22.2	70.8	48.5	66.6	44.4	10	0	63.9	25.8	32.2	17.8	0	25.3	0	0	14.7	37	55.5	20.5	58.2	18	20	17.9
3D-SIS2018.12		26.8	23.2	78.6	47.7	63.3	37	19.6	0	0	21.3	34.4	16.8	0	16.7	0	10	22.8	59.7	49.2	10	77.2	10	0	19.3
SGPN2018.12		22.5	18.6	39.2	28.5	46.5	26.7	21.8	15.9	0	24.9	23.9	16.3	20.8	15.1	10.7	0	17.7	35.1	37	22.9	34.2	17.7	31.5	13.9
Mask R-CNN2018.12		19.9	14.9	19	19.5	13.5	12.2	11.7	14.2	35	15.7	18.3	13.7	0	24.4	23.1	26	28.8	51.2	28.1	14.7	32.2	11.4	10.7	19.5
Seg-Cluster2018.12		16.4	16.8	16.2	15.6	11.8	14.5	10	11.7	27.2	20	25.7	10	0	15	0	20	27.8	39.5	22.9	10.7	38.9	10.4	0	12.3