AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

CoDA: Color Distribution Probing for Efficient and Generalizable AI-Generated Image Detection

Zexi Jia, Zhiqiang Yuan, Xiaoyue Duan, Jinchao Zhang, Jie Zhou, Anil K. Jain

TPAMI 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 the color branch (Noise-Quantization Probe difference map) to the image branch (ablation)CoDA image branch · FakeForm 62 domainsaccuracy (%): 66.8 -> 77.7Improved
add the image branch to the color branch (ablation)CoDA color branch · FakeForm 62 domainsaccuracy (%): 58.4 -> 77.7Improved
add the color branch (ablation)CoDA image branch · ForenSynths (GAN)accuracy (%): 95.5 -> 98.7Improved
add the color branch (ablation)CoDA image branch · GenImageaccuracy (%): 95.1 -> 95.9Improved
add the color branch (ablation)CoDA image branch · FakeForm Gamesaccuracy (%): 77.0 -> 92.2Improved
add the color branch (ablation)CoDA image branch · FakeForm Cartoon (B&W)accuracy (%): 66.9 -> 68.2Improved
increase noise replicas R from 20 to 50 (ablation)CoDA · FakeForm 62 domainsaccuracy (%): 75.7 -> 77.7Improved
raise noise level sigma from 0.05 to 0.10 (ablation)CoDA · FakeForm 62 domainsaccuracy (%): 75.1 -> 77.7Improved
raise noise level sigma from 0.10 to 0.20 (ablation)CoDA · FakeForm 62 domainsaccuracy (%): 77.7 -> 75.3Hurt
replace CLIP backbone with ResNet (both with color branch) (ablation)CoDA · FakeForm 62 domainsaccuracy (%): 76.9 -> 77.7Improved
add the color branch to CLIP (ablation)CLIP Only · FakeForm 62 domainsaccuracy (%): 75.2 -> 76.9Improved
apply JPEG compression QF=75 to test imagesCoDA · GenImage-style diffusion testaccuracy (%): 95.3 -> 93.7Hurt

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