AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Fighting Fake News: Image Splice Detection via Learned Self-Consistency

Minyoung Huh, Andrew Liu, Andrew Owens, Alexei A. Efros

ECCV 2018 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Pretext-task visual pre-training · 481 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 EXIF metadata consistency instead of directly predicting same-image consistency (ablation)patch consistency model · Splice detection, CarvalhomAP: 0.75 -> 0.87Improved
use EXIF metadata consistency instead of same-image consistency (ablation)patch consistency model · Splice localization, HaysmAP: 0.21 -> 0.48Improved
use EXIF metadata consistency instead of same-image consistency (ablation)patch consistency model · Splice localization, CarvalhomAP: 0.36 -> 0.51Improved
use EXIF metadata consistency instead of same-image consistency (ablation)patch consistency model · Splice localization, Realistic TamperingmAP: 0.21 -> 0.20No clear effect
train longer on EXIF vs Image consistency (ablation)Image-Consistency · splice localizationgap between modelsMixed
use self-supervised EXIF consistency instead of a supervised FCN trained on spliced imagespatch consistency model · Splice localization, ColumbiamAP: 0.60 -> 0.91Improved
use EXIF consistency instead of camera classification (ablation)patch consistency model · Splice detection, Realistic TamperingmAP: 0.15 -> 0.55Improved
use EXIF consistency instead of E-MFCNpatch consistency model · Splice localization vs Salloum et al., Columbia (full)F1: 0.61 -> 0.88Improved
use EXIF consistency instead of E-MFCNpatch consistency model · Splice localization vs Salloum et al., Carvalho (full)F1: 0.48 -> 0.52Improved
use EXIF consistency instead of CFApatch consistency model · Splice localization, Realistic TamperingmAP: 0.40 -> 0.20Hurt
add post-processing consistency (re-JPEG, blur, resize) tasks (ablation)EXIF-Consistency · Realistic Tampering / copy-movedetection of post-processing cuesImproved

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