Few-shot anomaly detection with adaptive feature transformation and descriptor construction  

作　　者：Zhengnan HU Xiangrui ZENG Yiqun LI Zhouping YIN Erli MENG Leyan ZHU Xianghao KONG 

机构地区：[1]School of Mechanical Science&Engineering,Huazhong University of Science and Technology,Hubei,Wuhan 430074,China [2]Xiaomi Corporation,Beijing 100085,China [3]Institute of Artificial Intelligence,Beihang University,Beijing 100191,China

出　　处：《Chinese Journal of Aeronautics》2025年第3期491-504,共14页中国航空学报(英文版)

基　　金：supported by the National Natural Science Foundation of China(No.52188102).

摘　　要：Anomaly Detection (AD) has been extensively adopted in industrial settings to facilitate quality control of products. It is critical to industrial production, especially to areas such as aircraft manufacturing, which require strict part qualification rates. Although being more efficient and practical, few-shot AD has not been well explored. The existing AD methods only extract features in a single frequency while defects exist in multiple frequency domains. Moreover, current methods have not fully leveraged the few-shot support samples to extract input-related normal patterns. To address these issues, we propose an industrial few-shot AD method, Feature Extender for Anomaly Detection (FEAD), which extracts normal patterns in multiple frequency domains from few-shot samples under the guidance of the input sample. Firstly, to achieve better coverage of normal patterns in the input sample, we introduce a Sample-Conditioned Transformation Module (SCTM), which transforms support features under the guidance of the input sample to obtain extra normal patterns. Secondly, to effectively distinguish and localize anomaly patterns in multiple frequency domains, we devise an Adaptive Descriptor Construction Module (ADCM) to build and select pattern descriptors in a series of frequencies adaptively. Finally, an auxiliary task for SCTM is designed to ensure the diversity of transformations and include more normal patterns into support features. Extensive experiments on two widely used industrial AD datasets (MVTec-AD and VisA) demonstrate the effectiveness of the proposed FEAD.

关 键 词：Industrial applications Anomaly detection Learning algorithms Feature extraction Feature selection 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP181[自动化与计算机技术—计算机科学与技术]