Fake Image Detection on AIGCDetectBench (test) benchmark leaderboard · SOTA2 Research

Fake Image Detection on AIGCDetectBench (test)

99.8Reference AccuracyLadeda

Evaluation Results

Method	Links
Ladeda2025.08		99.8	99.7	7.2	29	95.6	98.3	4.9	0	19.2	0.1	100	99.9	27.3	79.7	88.8	97.9	15.5	92.4	58.6
C2P-CLIP2025.08		99.8	100	99.8	99.8	72.2	78	87.7	100	34.9	67.6	100	99.9	50.5	73.3	28.4	99.7	93.3	98.1	82.4
Effort2025.08		99.8	100	100	100	79.6	82.1	100	100	91.1	76.6	100	99.9	53.1	81.2	44.2	81.2	96.3	75.8	87.8
NPR2025.08		99.3	98.9	29.3	16.5	67.1	58.7	1.7	6.5	7.9	0.1	100	99.9	26.5	69.2	71	97.7	15.4	89.8	53.1
SAFE2025.08		99.2	99.7	85.6	83.5	88.1	96.7	89	99.9	8.4	17.3	99.8	99.6	36.8	90.5	86.3	98.5	84	92	80.8
CNN-Spot2025.08		99	95.3	18.7	1.8	36.5	22	2.5	23.1	1.1	29.3	55.9	55.6	1.8	4.8	5.2	27.6	0.7	4.5	29
DFFreq2025.08		98	98	70.9	93.6	99	98.9	93.8	97.4	3.5	76	99.9	99.8	65.9	87.8	94.8	99.1	76	95.9	85.2
NS-Net2025.08		97.7	100	99.9	100	95.7	98.6	98.5	100	68.9	60.6	100	99.9	85.3	97.2	77.4	100	98.9	98.8	93.2
VIB-Net2025.08		93.5	99.4	97.3	91.3	71.3	73.3	97.8	97.3	97.3	90.6	100	99.8	52.8	69.3	63.7	99.3	80.9	58.5	84.7
AIDE2025.08		93	95.3	88	96.9	89.6	97	90	97	42.9	9.2	99.8	99.7	92.4	98.7	63.7	98.9	90.1	98.9	85.6
UnivFD2025.08		92.3	98.9	90.5	79.2	55.7	48.7	91.1	96.9	92.8	26.9	96.3	96	12.7	75.6	61.2	84.7	45.6	62.3	72.7
FreqNet2025.08		89.9	99.4	57.1	51	75.1	67.5	9.9	88.4	95.4	35.8	99.9	99.8	37.7	78.9	80.8	98	34.1	88.8	71.7