Adversarial attack
In AI detection contexts, a forgery specifically constructed to defeat a trained detector by incorporating small perturbations that cause the detector's classifier to output a 'genuine' prediction. A known limitation of neural-network-based forgery detectors operating against motivated, technically sophisticated adversaries.
Explained in
Related terms
Automatic Signature Verification (ASV)C2PA (Coalition for Content Provenance and Authenticity)Content-aware fillDiffusion modelESDA (Electrostatic Detection Apparatus)GAN (Generative Adversarial Network)GAN spectrum fingerprintJUMBF (JPEG Universal Metadata Box Format)PRNU (Photo Response Non-Uniformity)