Get Started
Home
Topics
Search
Library
Research questionHow can attackers evade black-box AIGC detectors using a frozen diffusion model without source images or detector-aware retraining?AIGC detectors must identify synthetic images even when generation can be adjusted to reduce detectable signals. This is difficult when the attacker can observe only detector responses and cannot alter the generator or rely on source images.
AI
Computer Vision
Diffusion Models
Image & Video Processing
Image Generation
Latest papersRecent research connected to this question, newest first.TIGA: Trajectory-Injected Generative Attack against Black-box AIGC DetectorsThe evidence concerns diffusion-generated facial images and specialized forensic detectors accessed through black-box queries. Reported results cover transfer to unseen detectors, perceptual quality, and robustness to common post-processing; the input does not establish broader deployment conditions.research paper · Sep 2, 2026
Related questions
How can deployed vision models resist diverse backdoor triggers without model internals or auxiliary data?How can black-box adversarial attacks disrupt image-based continuous-control agents with fewer environment queries?How should diffusion samplers allocate limited neural evaluations without sacrificing generated image quality?How can frozen dense-prediction models remain robust under distribution shift without retraining or changing their prediction heads?