AI Paper Atlas

Research map / Image generation / Autoregressive and latent generation

Hallucination Early Detection in Diffusion Models

Federico Betti, Lorenzo Baraldi, Lorenzo Baraldi, Rita Cucchiara, Nicu Sebe

IJCV 2026 · Image generation › Autoregressive and latent generation · approach: Reconstruction-based diffusion 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 HEaD+ to SD 1.4SD 1.4 · COCO multi-objectMG3: 50.74 -> 65.03Improved
add HEaD+ to TokenComposeTokenCompose · COCO multi-objectMG3-MG5: 76.16 -> 81.27 MG3, 28.81 -> 35.33 MG4, 3.28 -> 4.93 MG5Improved
add HEaD+ to TokenComposeTokenCompose · COCO multi-objectMG2: 98.08 -> 97.61No clear effect
add HEaD+ to PixArt-alphaPixArt-alpha · COCO multi-objectMG2-MG5: 83.19 -> 89.19 MG3, 42.80 -> 51.65 MG4Improved
use earlier detection timestep T=5 instead of T=25 (ablation)HEaD+ · InsideGen time savingtime saved: 14.18 -> 37.73Improved
use later detection timestep T=25 instead of T=5 (ablation)HEaD+ · InsideGenTN-Rate: 50.16 -> 76.95Improved
use HEaD+ architecture instead of HEaDHEaD · InsideGen time saving (T=25)time saved: -11.70 -> 14.18Improved
use HEaD+ instead of HEaD at T=8HEaD · InsideGen time saving (T=8)time saved: 11.55 -> 37.73Improved
add textual token v_o as input (ablation)HEaD+ HP network · InsideGen (T=25)TN-Rate: 52.16 -> 76.95Improved
add PFI as input (ablation)HEaD+ HP network · InsideGen (T=25)TN-Rate: 62.35 -> 76.95Improved
add cross-attention maps as input (ablation)HEaD+ HP network · InsideGen (T=25)TN-Rate / Recall: 74.84 -> 76.95 TN-Rate, 91.21 -> 93.40 RecallMixed
add localization module to HEaD+ (TokenCompose)TokenCompose w/ HEaD+ · two-object relation promptsRelation Consistency: 50.0 -> 67.5Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

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