AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

CHIMERA: Controllable High-quality Image-Mask Extraction for Reliable Diffusion-based Anomaly Synthesis

JoungBin Lee, Hyunkoo Lee, Jini Yang, Chaehyun Kim, Jung Yi, Seok Hwangbo, Hyeoncheol Lee, Minho Chun, Eunjo Jeong, Seungryong Kim

AAAI 2026 · Image generation › Text-to-image diffusion · approach: Compositional text-to-image guidance · 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 focal mask loss on attention maps (ablation)LoRA-only fine-tuning · MVTec AD ablationPRO: 47.6 -> 90.4Improved
add focal mask loss on attention maps (ablation)LoRA-only fine-tuning · MVTec AD ablationAUC-P: 74.4 -> 96.8Improved
add mask decoder on top of mask loss (ablation)LoRA + mask loss · MVTec AD ablationF1-P: 64.9 -> 71.2Improved
add layer selection on top of decoder (ablation)LoRA + mask loss + decoder · MVTec AD ablationPRO: 90.4 -> 88.6No clear effect
use generic 'sks' token instead of anomaly-type text (ablation)CHIMERA · MVTec AD ablationPRO: 96.1 -> 37.4Hurt
omit spatial location from prompt (ablation)CHIMERA · MVTec AD ablationPRO: 96.1 -> 81.6Hurt
omit spatial location from prompt (ablation)CHIMERA · MVTec AD ablationF1-P: 75.0 -> 54.1Hurt
use structured VLM captions with SD3 and LoRA vs prior generatorsSD3 CHIMERA · MVTec AD generationInception Score: 2.02 -> 2.40Improved
use CHIMERA vs AnomalyAnythingAnomalyAnything · MVTec AD generationIC-LPIPS: 0.33 -> 0.27Hurt
use CHIMERA unified model vs per-class DRAEMDRAEM · MVTec AD 1-shot detectionAUC-I: 94.6 -> 94.61No clear effect
use unified CHIMERA vs per-class AnoDiffAnoDiff · MVTec AD 1-shot detectionAUC-I: 99.2 -> 94.61Hurt
use CHIMERA on unseen mobile panel data vs AnomalyAnythingAnomalyAnything · Mobile device panel 1-shotAUC-I: 80.7 -> 84.1Improved

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