AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Pixel-Inconsistency Modeling for Image Manipulation Localization

Chenqi Kong, Anwei Luo, Shiqi Wang, Haoliang Li, Anderson De Rezende Rocha, Alex Chichung Kot

TPAMI 2025 · Image generation › Autoregressive and latent generation · approach: Generalizable fake image detection · 92 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 boundary decoder (ablation)PIM backbone · 11 unseen datasets (trained on CASIAv2), averageavg F1: .211 -> .220Improved
add regular data augmentation (RDA) (ablation)PIM backbone + boundary decoder · 11 unseen datasets (trained on CASIAv2), averageavg F1: .220 -> .233Improved
add PIDA instead of RDA (ablation)PIM backbone + boundary decoder · 11 unseen datasets (trained on CASIAv2), averageavg F1: .233 -> .260Improved
add PIDA on top of RDA (ablation)PIM backbone · 11 unseen datasets (trained on CASIAv2), averageavg F1: .260 -> .283Improved
add central difference convolution (CDC) (ablation)PIM · 11 unseen datasets (trained on CASIAv2), averageavg F1: .283 -> .304Improved
add radial difference convolution (RDC) alongside CDC (ablation)PIM · 11 unseen datasets (trained on CASIAv2), averageavg F1: .304 -> .312Improved
add Learning-to-Weight Module (ablation)PIM · 11 unseen datasets (trained on CASIAv2), averageavg F1: .312 -> .317Improved
add compactness loss (ablation)PIM · 11 unseen datasets (trained on CASIAv2), averageavg IoU: .271 -> .269No clear effect
add Global Pixel Dependency Encoder (masked attention) (ablation)PIM · 11 unseen datasets (trained on CASIAv2), averageavg F1: .323 -> .330Improved
add image reconstruction loss (ablation)PIM · 11 unseen datasets (trained on CASIAv2), averageavg F1: .330 -> .333Improved
multi-head instead of single-head attention in global encoder (ablation)PIM · 11 unseen datasets (trained on CASIAv2), averageavg F1: .312 -> .333Improved
multi-head instead of single-head attention in local encoder (ablation)PIM · 11 unseen datasets (trained on CASIAv2), averageavg F1: .318 -> .333Improved

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