Quality-Aware Language-Conditioned Local Auto-Regressive Anomaly Synthesis and Detection

Despite substantial progress in anomaly synthesis methods, existing diffusion-based and coarse inpainting pipelines commonly suffer from structural deficiencies such as microstructural discontinuities, limited semantic controllability, and inefficient generation. To overcome these limitations, we introduce ARAS, a language-conditioned, autoregressive anomaly synthesis approach that precisely injects local, text-specified defects into normal images via tokenanchored latent editing. Leveraging a hard-gated autoregressive operator and a training-free, context-preserving masked sampling kernel, ARAS significantly enhances defect realism, preserves fine-grained material textures, and provides continuous semantic control over synthesized anomalies. Integrated within our Quality-Aware Reweighted Anomaly Detection (QARAD) framework, we further propose a dynamic weighting strategy that emphasizes high-quality synthetic samples by computing an image-text similarity score with a dual-encoder model. Extensive experiments across three benchmark datasets-MVTec AD, VisA, and BTAD, demonstrate that our QARAD outperforms SOTA methods in both image-and pixel-level anomaly detection tasks, achieving improved accuracy, robustness, and a 5× synthesis speedup compared to diffusion-based alternatives. Our complete code and synthesized dataset will be publicly available.