List your company
ResearchFormerly WizWand

AI-Generated Image Detection on WildRF (In-the-wild)

98.8Accuracy (Real)LaDeDa
14.66436.50758.3580.193Feb 2, 2026
Updated 8mo ago

Evaluation results

MethodLinks
LaDeDaEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0298.811.955.4
NPREvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.029824.361.2
AIDEEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0297.319.558.4
CNNSpotEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0295.92962.5
DINOv3-LinearEvaluation Protocol=Linear Probe, Training Dataset=GenImage SD v1.42026.0294.897.596.1
Gram-NetEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0294.720.557.6
SAFEEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0291.830.961.3
DDAEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0289.990.890.4
FreqNetEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0273.155.964.5
UnivFDEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0269.363.766.5
PE-CLIP-LinearEvaluation Protocol=Linear Probe, Training Dataset=GenImage SD v1.42026.0267.999.483.6
DINOv2-LinearEvaluation Protocol=Linear Probe, Training Dataset=GenImage SD v1.42026.0264.377.270.5
OMATEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0263.371.567.4
SigLIP2-LinearEvaluation Protocol=Linear Probe, Training Dataset=GenImage SD v1.42026.0259.798.479
MetaCLIP2-LinearEvaluation Protocol=Linear Probe, Training Dataset=GenImage SD v1.42026.0247.897.972.8
MetaCLIP-LinearEvaluation Protocol=Linear Probe, Training Dataset=GenImage SD v1.42026.0246.192.369.2
SigLIP-LinearEvaluation Protocol=Linear Probe, Training Dataset=GenImage SD v1.42026.0238.389.764
EffortEvaluation Protocol=Standard Detection, Training Dataset=GenImage SD v1.42026.0217.995.556.7