[AAAI 2026] Anomagic: Crossmodal Prompt-driven Zero-shot Anomaly Generation

Yuxin Jiang1, Wei Luo3, Hui Zhang2, Qiyu Chen4, Haiming Yao3, Weiming Shen1,†, Yunkang Cao2,*
1Huazhong University of Science and Technology, 2Hunan University, 3Tsinghua University, 4Institute of Automation, Chinese Academy of Sciences
†Co-corresponding author    *Corresponding author
Project Page Paper Code HuggingFace Demo (coming soon)

Abstract

We propose Anomagic, a zero-shot anomaly generation method that produces semantically coherent anomalies without requiring any exemplar anomalies. By unifying both visual and textual cues through a crossmodal prompt encoding scheme, Anomagic leverages rich contextual information to steer an inpainting‐based generation pipeline. A subsequent contrastive refinement strategy enforces precise alignment between synthesized anomalies and their masks, thereby bolstering downstream anomaly detection accuracy.

To facilitate training, we introduce AnomagicDataset, a collection of 12,987 anomaly–mask–caption triplets assembled from 13 publicly available datasets, where captions are automatically generated by multimodal large language models using structured visual prompts and template‐based textual hints. Extensive experiments demonstrate that Anomagic trained on AnomagicDataset can synthesize more realistic and varied anomalies than prior methods, yielding superior improvements in downstream anomaly detection. Furthermore, Anomagic can generate anomalies for any normal‐category image using user‐defined prompts, establishing a versatile foundation model for anomaly generation.

Method

Overall Framework of Anomagic

The crossmodal prompt-driven zero-shot anomaly generation pipeline with contrastive refinement.

Anomagic Framework

Zero-shot Generation Results under Prompts in AnomVerse

Zero-shot Generation Results under User Defined Prompts

BibTeX

@article{anomagic2025,
  title={Anomagic: Crossmodal Prompt-driven Zero-shot Anomaly Generation},
  author={Jiang, Yuxin and Luo, Wei and Zhang, Hui and Chen, Qiyu and Yao, Haiming and Shen, Weiming and Cao, Yunkang},
  journal={arXiv preprint arXiv:2511.10020},
  year={2025}
}