Segmentation on SAT-MTB v1 (test) benchmark leaderboard · SOTA2 Research

Segmentation on SAT-MTB v1 (test)

35.9mAP50PointRend

Evaluation Results

Method	Links
PointRend2026.06		35.9	2.1	67.1	11.8	54.6	45.9	57.4	0	35	70.3	6.9	47.5	32.5
SCNet2026.06		35.9	18.6	85.1	2	66.3	38.5	56.2	0	40.2	73.2	0	22.5	28.3
HTC2026.06		33.3	5.7	80.2	1.3	38.5	36.5	62.2	0	43.9	72.1	0	30.5	28.6
HTC2026.06		32	16.1	76.1	5.7	26.4	36.4	51.2	0	40.4	71.1	0.1	30.5	30.5
Cascade Mask RCNN2026.06		31.8	22.9	77.8	0.2	64.3	32.7	55.4	0	26.9	53.1	0	26.3	21.8
DetectorRS2026.06		31.4	11.5	83.5	3.5	10.8	33.8	52.9	0	55.4	76.7	0.9	23.9	24.5
Mask RCNN2026.06		31.2	0	64.5	3.7	36.8	45.2	56.3	0	31.9	61.3	4.1	45.9	24.2
SCNet2026.06		31.2	9	89.3	1.5	0.3	38.7	66.3	0	55.1	74.5	0.1	28.8	11
Cascade Mask RCNN2026.06		29.3	4.3	70	0.8	30.6	38.8	53	0	37.4	60.1	0.2	24	32.6
PointRend2026.06		27.8	9.8	61.6	26.6	2.7	31.8	55.3	0	37.8	64.1	0	23.4	20.4
MaskScoring RCNN2026.06		27.7	3.6	73.8	3.2	8.3	41.1	48.4	0	39.8	71.5	0.2	1.8	24.5
Mask RCNN2026.06		26.6	4.8	60.5	1.6	15.5	36.9	52.7	0	40.8	66.9	0	24.8	14.1
MTTR2026.06		19.4	1.2	53.1	7.6	3.6	21.6	31.8	0	26.5	23.1	0.2	9.4	19.8
Mask Transfiner2026.06		17.6	0.9	48.5	4.6	1.6	13.5	24.8	0	13.4	41.6	0	7.6	15.4
YOLACT2026.06		13.6	0	19.5	0	0	26.4	11.3	0	37.4	42.6	0.6	4.9	21
QueryInst2026.06		12.9	0.5	54.2	0	0.3	14.5	22.5	0	11.3	16.3	1.4	10.5	23.1
IDOL2026.06		10.5	0	24.6	1.3	2.8	10.2	2.8	0	16.5	13	0	0.2	12.6
SeqFormer2026.06		10.3	0	29.8	0.6	3.8	0.6	12.7	0	21.1	16.2	0	1.6	20.9