Table 4 Performance Metrics for each Dataset when tested on ResNet50 model trained with Stable Diffusion

From: Detection of AI generated images using combined uncertainty measures and particle swarm optimised rejection mechanism

Dataset

Accuracy

Precision

Recall

F1 Score

Midjourney

0.6602

0.5957

0.9972

0.7458

GLIDE

0.7361

0.6550

0.9977

0.7908

VQDM

0.5717

0.5387

0.9965

0.6994

BigGAN

0.5597

0.5319

0.9975

0.6938

StyleGAN3

0.4534

0.4534

1.0000

0.6239