AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

ForensicConcept: Transferable Forensic Concepts for AIGI Detection

Menyanshu Zhou, Ziyin Zhou, Ke Sun, Yunpeng Luo, Jiayi Ji, Xiaoshuai Sun, Rongrong Ji

ICML 2026 · Image generation › Autoregressive and latent generation · approach: Generalizable fake image 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 concept-aligned projection (concept branch) to DINOv3 detector (ablation)DINOv3 detector · mean of GenImage, GAN-family, Chameleonaccuracy (%): 87.2 -> 88.8Improved
add concept branch to DINOv3 on GAN-family (ablation)DINOv3 detector · GAN-familyaccuracy (%): 87.3 -> 90.1Improved
add concept branch to DINOv3 on Chameleon (ablation)DINOv3 detector · Chameleonaccuracy (%): 83.7 -> 84.4Improved
add concept branch to DINOv3 on GenImage (ablation)DINOv3 detector · GenImage meanaccuracy (%): 90.7 -> 92.0Improved
inject CleanDIFT diffusion codebook into CLIP (ablation)CLIP detector · GenImage meanaccuracy (%): 83.7 -> 88.2Improved
inject CleanDIFT codebook into CLIP on Midjourney (ablation)CLIP detector · GenImage Midjourneyaccuracy (%): 70.4 -> 85.9Improved
inject CleanDIFT codebook into CLIP on SDv1.4, SDv1.5, Wukong (ablation)CLIP detector · GenImage SDv1.4 / SDv1.5 / Wukongaccuracy (%): -0.9, -1.1, -0.6Hurt
use diffusion-derived CleanDIFT codebook rather than other backbone codebooks for injection into CLIP (ablation)CLIP detector · GenImage meanaccuracy (%): 88.2 vs 87.5 (EfficientNet), 85.4 (DINOv3)Improved
use more CKNNA-aligned backbone as codebook source (ablation)CLIP codebook injection · GenImage meanaccuracy (%)Mixed
increase CKNNA neighborhood size kNN from 5 to 40 (ablation)CKNNA alignment analysis · CleanDIFT us6CKNNAHurt
use ForensicConcept DINOv3 detector instead of CLIP or Effort under JPEG QF=75AIGI detector · GenImage with post-processingaccuracy (%): 63.37 -> 91.83 (vs Effort)Improved

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