ACD-CLIP: Decoupling Representation and Dynamic Fusion for Zero-Shot Anomaly Detection

ABSTRACT Pre-trained Vision-Language Models (VLMs) struggle with Zero-Shot Anomaly Detection (ZSAD) due to a critical adaptation gap: they lack the local inductive biases required for dense prediction and employ inflexible feature fusion paradigms. We address these limitations through an Architectural Co-Design framework that jointly refines feature representation and cross-modal fusion. Our method proposes a parameter-efficient Convolutional Low-Rank Adaptation (Conv-LoRA) adapter to inject local inductive biases for fine-grained representation, and introduces a Dynamic Fusion Gateway (DFG) that leverages visual context to adaptively modulate text prompts, enabling a powerful bidirectional fusion. Extensive experiments on diverse industrial and medical benchmarks demonstrate superior accuracy and robustness, validating that this synergistic co-design is critical for robustly adapting foundation models to dense perception tasks. Index T erms-- anomaly detection, multimodal feature fusion, vision-language model, transfer learning, PEFT 1. INTRODUCTION Zero-Shot Anomaly Detection (ZSAD) adapts Vision-Language Models (VLMs) [1, 2] like CLIP [3] to circumvent the extensive training data required by traditional methods [4, 5].