SSVP: Synergistic Semantic-Visual Prompting for Industrial Zero-Shot Anomaly Detection

TL;DR

SSVP introduces a novel approach combining semantic and visual prompts with hierarchical synergy and cross-modal attention to significantly improve zero-shot industrial anomaly detection performance.

Contribution

The paper presents SSVP, a new method that fuses diverse visual encodings and employs a dual-gated calibration to enhance zero-shot anomaly detection in industrial settings.

Findings

Achieves 93.0% Image-AUROC on MVTec-AD

Outperforms existing zero-shot methods significantly

Validates robustness across seven benchmarks

Abstract

Zero-Shot Anomaly Detection (ZSAD) leverages Vision-Language Models (VLMs) to enable supervision-free industrial inspection. However, existing ZSAD paradigms are constrained by single visual backbones, which struggle to balance global semantic generalization with fine-grained structural discriminability. To bridge this gap, we propose Synergistic Semantic-Visual Prompting (SSVP), that efficiently fuses diverse visual encodings to elevate model's fine-grained perception. Specifically, SSVP introduces the Hierarchical Semantic-Visual Synergy (HSVS) mechanism, which deeply integrates DINOv3's multi-scale structural priors into the CLIP semantic space. Subsequently, the Vision-Conditioned Prompt Generator (VCPG) employs cross-modal attention to guide dynamic prompt generation, enabling linguistic queries to precisely anchor to specific anomaly patterns. Furthermore, to address the discrepancy between global scoring and local evidence, the Visual-Text Anomaly Mapper (VTAM) establishes a dual-gated calibration paradigm. Extensive evaluations on seven industrial benchmarks validate the robustness of our method; SSVP achieves state-of-the-art performance with 93.0% Image-AUROC and 92.2% Pixel-AUROC on MVTec-AD, significantly outperforming existing zero-shot approaches.