AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Learning on Gradients: Generalized Artifacts Representation for GAN-Generated Images Detection

Chuangchuang Tan, Yao Zhao, Shikui Wei, Guanghua Gu, Yunchao Wei

CVPR 2023 · Image generation › Autoregressive and latent generation · approach: Generalizable fake image detection · 191 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
represent images with gradients from a pretrained StyleGAN-bedroom discriminator rather than raw images (ablation)LGrad · Cross-model & category, 1-class (mean)accuracy: 52.5 -> 86.3Improved
use gradients from a pretrained GAN discriminatorLGrad · Cross-model & category, 1-class (mean)accuracy: 74.9 -> 86.3Improved
use a pretrained ProGAN discriminator rather than a randomly initialized one (ablation)LGrad, CelebA-HQ to StyleGAN2 · CelebA-HQ to StyleGAN2accuracy: 52.4 -> 96.5Improved
use a GAN discriminator rather than a classification network as transformation model (ablation)LGrad · Cross-model & category, 1-class (mean)accuracy: 76.0 -> 86.3Improved
use InceptionV3 as the transformation model instead of StyleGAN-bedroom (ablation)LGrad · Cross-model & category, 1-class (mean)accuracy: 86.3 -> 56.9Hurt
increase training data from 4k to 36k images (ablation)LGrad with StyleGAN-bedroom transformation · Cross-model & category (mean)accuracy: 75.9 -> 86.3Improved
increase training data from 36k to 72k images (ablation)LGrad with StyleGAN-bedroom transformation · Cross-model & category (mean)accuracy: 86.3 -> 86.2No clear effect
blur test images (ablation)LGrad with StyleGAN-bedroom transformation · Cross-model & categorymean accuracy: 86.3 -> 78.5Hurt
crop test images (ablation)LGrad with StyleGAN-bedroom transformation · Cross-model & categorymean accuracy: 86.3 -> 85.0No clear effect
apply JPEG compression to test images (ablation)LGrad with StyleGAN-bedroom transformation · Cross-model & categorymean accuracy: 86.3 -> 69.5Hurt
add random noise to test images (ablation)LGrad with StyleGAN-bedroom transformation · Cross-model & categorymean accuracy: 86.3 -> 69.1Hurt
apply combined blur, cropping, JPEG, and noise perturbations (ablation)LGrad with StyleGAN-bedroom transformation · Cross-model & categorymean accuracy: 86.3 -> 75.9Hurt

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