Anomaly Detection on MAD benchmark leaderboard · SOTA2 Research

Anomaly Detection on MAD

93.6GorillaOmniposeAD

Evaluation Results

Method	Links
OmniposeAD2023.10		93.6	94	84.7	95.6	82.5	92.4	88.2	95.7	86.5	90.1	88.3	88.2	82.3	92.5	97	84.9	91.5	97.1	98.8	93.5	90.9
CFlow2023.10		69.2	82.3	74.9	51	57	75.6	76.5	71.3	67.4	80.9	72.1	66.9	64.4	70.1	67.9	65.8	79.5	80	81.4	71.4	71.3
Patchcore2023.10		66.8	92.4	59.3	87	86	82.9	72.9	76.6	75.2	89.4	85.7	68.2	71.4	78.6	78	78.7	82.1	89.5	84.2	65.6	78.5
STFPM2023.10		65.3	79.6	42.2	64.4	64.1	52.4	72.7	56	53.6	56.5	50.6	53.7	56.7	61.7	61.1	52.2	68.9	60.8	60.7	56.7	59.5
DRAEM2023.10		58.9	70.4	34.5	18.4	65.8	69.1	67.2	68.6	59.7	59.5	64.4	51.7	53.1	62.5	62.3	61.3	83.7	73.9	76.1	57.4	60.9
UniAD2023.10		56.6	73	70	50.2	75.5	74.7	65.3	61.2	57.5	70.4	54.6	50.5	55.4	59.3	53.4	53.1	69.5	67.7	65.1	55.6	62.2
Fastflow2023.10		51.1	45	72.1	67.7	53.2	76.5	58.2	70.5	63.9	71.4	59.6	54.9	53.4	61.6	53.4	51.3	51.9	67.6	72.9	59.5	60.8
FAVAE2023.10		46.8	68.3	33.6	82.8	62.5	73.3	62.5	82.4	50.6	74.9	52.5	34.6	65.2	49.1	46.1	53.2	66.9	58.2	52.8	43.5	58
CFA2023.10		41.8	85.6	36.6	58.3	77.7	78.4	74	64.2	66.7	86.5	66.7	52.1	65.9	71.7	69.8	68.2	91.4	80.6	78.7	53.7	68.2