What does this research mean for the field?

A dynamic visual adaptation framework integrating a Hyper AD Plug-in and Mixture-of-Experts module achieves state-of-the-art accuracy in multi-class anomaly detection by dynamically adjusting model parameters to input data. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

November 8, 2025AI1 citationsOpen Access

DVAD: A Dynamic Visual Adaptation Framework for Multi-Class Anomaly Detection

HGHan GaoAsia University HLHuiyuan LuoChinese Academy of Sciences FSFei ShenSichuan Agricultural University

Key Points

Key points are not available for this paper at this time.

Abstract

Despite the superior performance of existing anomaly detection methods, they are often limited to single-class detection tasks, requiring separate models for each class. This constraint hinders their detection performance and deployment efficiency when applied to real-world multi-class data. In this paper, we propose a dynamic visual adaptation framework for multi-class anomaly detection, enabling the dynamic and adaptive capture of features based on multi-class data, thereby enhancing detection performance. Specifically, our method introduces a network plug-in, the Hyper AD Plug-in, which dynamically adjusts model parameters according to the input data to extract dynamic features. By leveraging the collaboration between the Mamba block, the CNN block, and the proposed Hyper AD Plug-in, we extract global, local, and dynamic features simultaneously. Furthermore, we incorporate the Mixture-of-Experts (MoE) module, which achieves a dynamic balance across different features through its dynamic routing mechanism and multi-expert collaboration. As a result, the proposed method achieves leading accuracy on the MVTec AD and VisA datasets, with image-level mAU-ROC scores of 98.8% and 95.1%, respectively.

问 AI

Bookmark

View Full Paper