A CLIP-feature normalizing-flow model trained on frequency-masked natural images, not on AI-generated images, detects images from unseen generators with 95.56% mAP on the UnivFD benchmark.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
CLIP-Flow: A Universal Discriminator for AI-Generated Images Inspired by Anomaly Detection
A CLIP-feature normalizing-flow model trained on frequency-masked natural images, not on AI-generated images, detects images from unseen generators with 95.56% mAP on the UnivFD benchmark.