Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Generative Image Detection on ForenSynths (test) (Robustness Evaluation)
Loading...
97.22
Accuracy (Clean)
Effort
74.0176
80.0413
86.065
92.0887
Apr 14, 2026
Accuracy (Clean)
Accuracy (JPEG Q=50)
Accuracy (Noise k=7 σ=2)
Accuracy (Blur v=25)
Average Accuracy
Updated 4d ago
Evaluation Results
Method
Method
Links
Accuracy (Clean)
Accuracy (JPEG Q=50)
Accuracy (Noise k=7 σ=2)
Accuracy (Blur v=25)
Average Accuracy
Effort
2026.04
97.22
60.32
68.56
90.88
79.24
LPT (Ours)
Foundation Model=EVA-CLIP
2026.04
96.01
90.38
93.67
95.23
93.82
RINE
2026.04
95.4
81.5
84.6
91.7
88.3
NPR
2026.04
91.8
50.2
50
65.7
64.42
MLEP
2026.04
89.85
48.9
49.9
74.9
65.88
DDA
2026.04
84.16
78.02
79.65
83.4
81.3
CNN
2026.04
74.91
57.57
70.48
74.8
69.44
Feedback
Search any
task
Search any
task
