AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

LEGION: Learning to Ground and Explain for Synthetic Image Detection

Hengrui Kang, Siwei Wen, Zichen Wen, Junyan Ye, Weijia Li, Peilin Feng, Baichuan Zhou, Bin Wang, Dahua Lin, Linfeng Zhang, Conghui He

ICCV 2025 · Image generation › Autoregressive and latent generation · approach: Generalizable fake image detection · 49 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 LEGION MLLM instead of PAL4VST expert modelartifact localization · SynthScars ObjectF1: 19.25 -> 29.90Improved
use LEGION instead of PAL4VSTartifact localization · LOKI (unseen)mIoU: 52.55 -> 54.52 (Animal) ; 47.34 -> 48.66 (LOKI)Improved
use LEGION instead of LISA-v1-7Bartifact localization · RichHF-18K (unseen)F1: 21.94 -> 17.41Hurt
use LEGION instead of LLaVA-v1.6 for explanationsexplanation generation · SynthScarsROUGE-L: 29.61 -> 39.50Improved
use LEGION instead of LLaVA-v1.6 for explanationsexplanation generation · LOKIROUGE-L: 16.07 -> 18.55Improved
use LEGION instead of UniFD for detectiondetection (trained on ProGAN) · UniversalFakeDetect GANsaccuracy: 95.25 -> 97.01Improved
use LEGION instead of FreqNet for detectiondetection (trained on ProGAN) · UniversalFakeDetect Deepfakesaccuracy: 97.40 -> 63.37Hurt
add LEGION-guided prompt revision to regenerationSD3.5 regeneration · SynthScars test (200 images)HPS: 31.24 -> 33.36 (6.98%)Improved
add LEGION mask and explanation guidance to inpaintingSDXL inpainting · SynthScars test (200 images)HPS: 29.57 -> 30.20 (2.14%)Improved
use GPT-4o instead of LEGION for explanationsexplanation generation · SynthScarsCSS: 72.60 -> 53.55Hurt

Later papers that compared with this method

1 later accepted paper reports LEGION as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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