AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Leveraging Representations from Intermediate Encoder-blocks for Synthetic Image Detection

Christos Koutlis, Symeon Papadopoulos

ECCV 2024 · Image generation › Autoregressive and latent generation · approach: Generalizable fake image detection · 62 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
use intermediate CLIP block CLS tokens instead of last-layer features only (ablation)RINE · 20 test datasets, all generatorsaverage accuracy: 80.8 -> 89.8Improved
add Trainable Importance Estimator (TIE) (ablation)RINE · 20 test datasets, all generatorsaverage accuracy: 87.3 -> 89.8Improved
add supervised contrastive loss (ablation)RINE · 20 test datasets, all generatorsaverage accuracy: 88.4 -> 89.8Improved
add contrastive loss in 2-class setting (ablation)RINE 2-class · 20 test datasetsaccuracy: 87.9 -> 87.7Hurt
use CLIP ViT-L/14 instead of ViT-B/32 (ablation)RINE · 20 test datasetsACC and APImproved
vary contrastive loss factor xi (ablation)RINE · 20 test datasetsACC and APNo clear effect
train for more epochs (3 to 15) instead of 1 (ablation)RINE · 20 test datasetsACC and APHurt
increase training set size from 20% to 100% (ablation)RINE · 20 test datasetsACC and APNo clear effect
replace projections and TIE with Feature Pyramid Network fusion (ablation)RINE · 20 test datasets (4-class)average accuracy: 91.5 -> 69.6Hurt
use ImageNet ViT or Wang detector backbone instead of CLIP (ablation)RINE · 20 test datasets (4-class)average accuracy: 91.5 -> 60.1 (ViT); 91.5 -> 64.7 (Wang)Hurt
train on 20 classes instead of 4 (ablation)RINE · 20 test datasetsaverage accuracy: 91.5 -> 90.7No clear effect
apply combined perturbations at test timeRINE 1-class · 20 test datasetsaverage accuracy: 90.3 -> 80.6Hurt

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