English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Anomaly detectors are widely used in industrial production to detect andlocalize unknown defects in query images. These detectors are trained onnominal images and have shown success in distinguishing anomalies from mostnormal samples. However, hard-nominal examples are scattered and far apart frommost normalities, they are often mistaken for anomalies by existing anomalydetectors. To address this problem, we propose a simple yet efficient method:\textbf{H}ard Nominal \textbf{E}xample-aware \textbf{T}emplate \textbf{M}utual\textbf{M}atching (HETMM). Specifically, \textit{HETMM} aims to construct arobust prototype-based decision boundary, which can precisely distinguishbetween hard-nominal examples and anomalies, yielding fewer false-positive andmissed-detection rates. Moreover, \textit{HETMM} mutually explores theanomalies in two directions between queries and the template set, and thus itis capable to capture the logical anomalies. This is a significant advantageover most anomaly detectors that frequently fail to detect logical anomalies.Additionally, to meet the speed-accuracy demands, we further propose\textbf{P}ixel-level \textbf{T}emplate \textbf{S}election (PTS) to streamlinethe original template set. \textit{PTS} selects cluster centres andhard-nominal examples to form a tiny set, maintaining the original decisionboundaries. Comprehensive experiments on five real-world datasets demonstratethat our methods yield outperformance than existing advances under thereal-time inference speed. Furthermore, \textit{HETMM} can be hot-updated byinserting novel samples, which may promptly address some incremental learningissues.

Hard Nominal Example-aware Template Mutual Matching for Industrial Anomaly Detection