Semantic Occupancy Prediction on Occ-ScanNet (test) benchmark leaderboard · SOTA2 Research

Semantic Occupancy Prediction on Occ-ScanNet (test)

71.7Overall IoUFLM-Occ

Evaluation Results

Method	Links
FLM-Occ2026.06		71.7	62.5	70.6	58	53.1	58.8	74	77.9	63.5	55.2	70.8	60.3	64
FLM-Occ2026.06		71.2	61.9	70	57.5	52.6	58.1	73.5	77.5	62.7	55.5	70.3	59.8	63.6
FLM-Occ2026.06		65.3	55.1	64.5	52.2	43.3	48.3	66.5	70.6	53.5	45.6	64.4	50.7	55.9
FLM-Occ2026.06		64.7	55	64	52	44.7	48.6	67.3	70.4	54.3	46.7	64.3	52.4	56.3
SplatSSC2026.06		62.8	49.1	59	48.3	38.8	47.4	62.4	67	49.5	42.6	60.7	45.4	51.8
E.Occ++2026.06		54.9	36.4	53.1	41.8	34.4	42.9	57.3	64.1	45.2	34.8	54.2	44.1	46.2
E.Occ2026.06		53.6	39.6	50.4	41.4	31.7	40.9	55	61.4	44	36.1	53.9	42.2	45.2
DiScene2026.06		47.2	45.3	50.6	40.4	36.7	42.3	59.7	62	45.6	41.2	52.4	42.7	47.2
ISO2026.06		42.2	19.9	41.9	22.4	17	29	42.4	42	29.6	10.6	36.4	24.6	28.7
MonoScene2026.06		41.6	15.2	44.7	22.4	12.6	26.1	27	35.9	28.3	6.6	32.2	19.8	24.6