AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Your AI-Generated Image Detector Can Secretly Achieve SOTA Accuracy, If Calibrated

Muli Yang, Gabriel James Goenawan, Henan Wang, Huaiyuan Qin, Chenghao Xu, Yanhua Yang, Fen Fang, Ying Sun, Joo Hwee Lim, Hongyuan Zhu

AAAI 2026 · Image generation › Autoregressive and latent generation · approach: Generalizable fake image detection · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add supervised scalar logit calibration (100 target images)CNNSpot · AIGCDetectBenchmarkaverage accuracy: 70.83 -> 78.22Improved
add supervised scalar logit calibrationRINE (CLIP-based) · GenImageaverage accuracy: 81.78 -> 97.94Improved
add supervised scalar logit calibrationEffort · AIGCDetectBenchmarkaverage accuracy: 79.35 -> 89.48Improved
add unsupervised scalar logit calibration (no labels)RINE · AIGCDetectBenchmarkaverage accuracy: 86.77 -> 93.79Improved
add calibration to a detector already strong in-distributionAIDE · AIGCDetectBenchmarkaverage accuracy: 92.80 -> 94.42 (Sup.), 93.69 (Unsup.)Improved
add unsupervised calibration instead of supervisedFreDect · AIGCDetectBenchmarkaverage accuracy: 64.03 -> 65.09No clear effect
add calibration under JPEG compression (QF=90)AIDE · AIGCDetectBenchmark with JPEG QF=90average accuracy: 60.22 -> 75.61Improved
add calibration under Gaussian blur sigma=2.0LGrad · AIGCDetectBenchmark with blur sigma=2.0average accuracy: 50.03 -> 49.98No clear effect
use supervised KDE-based calibration vs unsupervised (ablation)Fusing · GenImageaverage accuracy: 80.94 -> 90.20Improved
reduce validation set to 10 images (ablation)supervised and unsupervised calibration · AIGCDetectBenchmark and GenImageaverage accuracyNo clear effect
estimate alpha with KDE instead of binary search or BCE training (ablation)supervised calibration of CNNSpot · AIGCDetectBenchmarkaverage accuracyImproved
add calibration to CNNSpot on GenImage (Unsup.)CNNSpot · GenImageaverage accuracy: 69.37 -> 70.86Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy