AI Paper Atlas

Research map / Detection and segmentation / CNN features

Decision-Driven Orthogonal Learning with Complementary Feature Mining for Robust Synthetic Image Detection

Kai Li, Wei Wang, Linchao Zhang, Siying Zhu, Wenqi Ren

AAAI 2026 · Detection and segmentation › CNN features · approach: Forgery detection · 0 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 decision-driven orthogonal constraint (ablation)Ours (ViT + conv branch) · ProGAN 2-class, quality-awareMean accuracy (%): 72.3 -> 75.4Improved
add adaptive global-local modulator (ablation)Ours without orthogonal constraint · ProGAN 2-class, quality-awareMean accuracy (%): 70.9 -> 72.3Improved
add bidirectional update strategy (ablation)Plain backbone · ProGAN 2-class, quality-awareMean accuracy (%): 63.5 -> 70.9Improved
replace gradient reversal compression removal with orthogonal constraintODDN · 2-class quality-awareMean accuracy (%): 71.4 -> 75.4Improved
same method in unknown-compression settingODDN · 2-class quality-agnosticMean accuracy (%): 70.7 -> 74.5Improved
same method with 4-class training dataODDN · 4-class quality-awareMean accuracy (%): 72.6 -> 76.5Improved
same method with 4-class training, unknown compressionODDN · 4-class quality-agnosticMean accuracy (%): 72.1 -> 75.5Improved
proposed method vs ODDN on AttGANODDN · 2-class quality-awareAccuracy (%): 71.2 -> 54.8Hurt
proposed method vs ODDN on BEGANODDN · 2-class quality-awareAccuracy (%): 42.3 -> 64.9Improved
proposed method vs ODDN on StarGANODDN · 2-class quality-awareAccuracy (%): 77.3 -> 92.3Improved
proposed method vs ODDN on RelGANODDN · 2-class quality-awareAccuracy (%): 72.5 -> 58.7Hurt
add DDOC and modules (logit separation)baseline detector · ProGAN and GauGAN testlogit distribution overlapImproved

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