AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

μFlow: Leveraging Average Images for Improving Generalisation of Deepfake Faces Detectors

Orazio Pontorno, Mattia Litrico, Luca Guarnera, Mario Valerio Giuffrida, Sebastiano Battiato

ECCV 2026 · Image generation › Autoregressive and latent generation · approach: Generalizable fake image detection

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
train only on real images instead of Real+GANs fake trainingμFlow vs ODDN · Real+GANs training, avg OOD accuracyaccuracy: 81.8 -> 96.8Improved
real-only training instead of Real+DM-CS trainingμFlow vs D3 · DM-OS test, accuracyaccuracy: +2.6Improved
real-only training instead of Real+DM-OS trainingμFlow vs D3 · Real+DM-OS training, avg accuracyaccuracy: 85.4 -> 93.5Improved
compare against detectors trained on mixed GAN and DM fakesμFlow vs D3 · Real+Mix*, avg accuracyaccuracy: 92.6 -> 94.7Improved
compare against real-only baselinesμFlow vs LAA-Net · Real-only baselines, avg accuracyaccuracy: 83.0 -> 94.7Improved
map features to average-image distribution instead of standard normal Gaussian (ablation)μFlow · WILD OODaccuracyImproved
use fixed threshold tau instead of LOF or OC-SVM (ablation)μFlow · WILD OODaccuracyNo clear effect
vary significance level alpha, choose alpha=0.01 (ablation)μFlow · WILD OODaccuracyNo clear effect
increase number K of averaged images from 500 to 700 (ablation)μFlow · Silhouette scoreSilhouette score: 0.72 -> 0.74Improved
test under resize, flip, Gaussian noise, salt-pepper without training on themμFlow vs D3 · content-preserving transformations (avg)accuracy: 89.4 -> 90.9Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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