Synthetic Image Detection on DiffusionForensics (test) benchmark leaderboard · SOTA2 Research

Synthetic Image Detection on DiffusionForensics (test)

93.5ADM AccuracyADOF

Evaluation Results

Method	Links
ADOF2026.06		93.5	99	99.6	100	99.2	100	99.9	100	97.4	99.9	97.1	99.8	99.8	100	99.9	100	98.3	99.8
NPR2026.06		88.6	98.9	99.8	100	91.8	99.8	100	100	91.2	100	100	100	97.4	99.8	93.8	100	95.3	99.8
LGrad2026.06		86.4	97.5	99.9	100	66.1	92.8	99.7	100	69.5	98.5	96.2	100	90.4	99.4	97.1	100	88.2	98.5
F3Net2026.06		80.9	96.9	84.7	99.4	74.7	98.9	100	100	72.8	99.5	100	100	73.4	97.2	99.8	100	85.8	99
Ojha2026.06		78.4	92.1	72.9	78.8	75	92.8	82.2	97.1	75.3	92.5	83.5	97.7	56.4	90.4	71.5	92.4	74.4	91.7
Patchfor2026.06		77.5	93.9	62.3	97.1	50	91.6	99.5	100	50.2	99.9	100	100	90.7	99.8	94.8	100	78.1	97.8
Frank2026.06		58.9	65.9	37	27.6	51.4	65	51.7	48.5	44	38.2	51.7	66.7	32.8	52.3	40.8	37.5	46	50.2
SelfBland2026.06		57	59	61.9	49.6	63.2	66.9	83.3	92.2	48.2	48.2	77.2	82.7	46.2	68	71.2	73.9	63.5	67.6
CNNDetection2026.06		53.9	71.8	62.7	76.6	50.2	82.7	50.4	78.7	50.8	90.3	50	71	38	76.7	52	90.3	51	79.8
GANDetection2026.06		51.1	53.1	62.3	46.4	50.2	63	51.6	48.1	50.6	79	51.1	51.2	39.8	65.6	50.1	36.9	50.8	55.4
Durall2026.06		39.8	42.1	52.9	49.8	55.3	56.7	43.1	39.9	44.5	47.3	38.6	38.3	39.5	56.3	62.1	55.8	47	48.3