Research map / Image generation / Autoregressive and latent generation
AnomalyPainter: Vision-Language-Diffusion Synergy for Realistic and Diverse Unseen Industrial Anomaly Synthesis
Zhangyu Lai, Yilin Lu, Xinyang Li, Jianghang Lin, Yansong Qu, Ming Li, Liujuan Cao
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add TALI (texture-aware latent init) to ADGM (ablation) | AnomalyPainter · VisA | IL / AUC-I: IL 0.25 -> 0.29; AUC-I 87.4 -> 91.6 | Improved |
| add ControlNet to ADGM (ablation) | AnomalyPainter · VisA | IL / AUC-I: IL 0.25 -> 0.28; AUC-I 87.4 -> 91.1 | Improved |
| combine TALI and ControlNet (ablation) | AnomalyPainter · VisA | IL / AUC-I: IL 0.33; AUC-I 92.6 (vs 91.6 / 91.1 single) | Improved |
| use VLLM-generated descriptions matched to Tex-9K textures instead of random textures (ablation) | AnomalyPainter · VisA | IL / AUC-I: IL 0.26 -> 0.33; AUC-I 90.9 -> 92.6 | Improved |
| start denoising at T*=16 instead of 20 or 12 (ablation) | AnomalyPainter · VisA | AUC-I: T*=20: 91.6; T*=12: 91.7; T*=16: 92.6 | Improved |
| AnomalyPainter data instead of RealNet data | 1-shot anomaly detection · VisA 1-shot | AUC-I: 86.0 -> 92.6 | Improved |
| AnomalyPainter instead of AnomalyAny data | 1-shot anomaly detection · MVTec AD 1-shot | PRO: AUC-I 94.9 -> 96.5; PRO 91.9 -> 92.5 | Mixed |
| AnomalyPainter instead of DRAEM data | 1-shot anomaly detection · VisA 1-shot | AUC-P: 97.5 -> 97.9 (within noise) | No clear effect |
Closest papers by meaning
- AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Models (AAAI 2024 · similarity 0.89)
- Unseen Visual Anomaly Generation (CVPR 2025 · similarity 0.88)
- Dual-Interrelated Diffusion Model for Few-Shot Anomaly Image Generation (CVPR 2025 · similarity 0.88)
- AnomalyDiffusion: Few-Shot Anomaly Image Generation with Diffusion Model (AAAI 2024 · similarity 0.88)
- FAST: Foreground‑aware Diffusion with Accelerated Sampling Trajectory for Segmentation‑oriented Anomaly Synthesis (NeurIPS 2025 · similarity 0.87)
- InCLAD: A Continual Learning Benchmark for Industrial Visual Anomaly Detection (NeurIPS 2026 · similarity 0.86)
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.
