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

Industrial anomaly detection (IAD) is crucial for automating industrialquality inspection. The diversity of the datasets is the foundation fordeveloping comprehensive IAD algorithms. Existing IAD datasets focus on thediversity of data categories, overlooking the diversity of domains within thesame data category. In this paper, to bridge this gap, we propose theAero-engine Blade Anomaly Detection (AeBAD) dataset, consisting of twosub-datasets: the single-blade dataset and the video anomaly detection datasetof blades. Compared to existing datasets, AeBAD has the following twocharacteristics: 1.) The target samples are not aligned and at differentscales. 2.) There is a domain shift between the distribution of normal samplesin the test set and the training set, where the domain shifts are mainly causedby the changes in illumination and view. Based on this dataset, we observe thatcurrent state-of-the-art (SOTA) IAD methods exhibit limitations when the domainof normal samples in the test set undergoes a shift. To address this issue, wepropose a novel method called masked multi-scale reconstruction (MMR), whichenhances the model's capacity to deduce causality among patches in normalsamples by a masked reconstruction task. MMR achieves superior performancecompared to SOTA methods on the AeBAD dataset. Furthermore, MMR achievescompetitive performance with SOTA methods to detect the anomalies of differenttypes on the MVTec AD dataset. Code and dataset are available athttps://github.com/zhangzilongc/MMR.

Industrial Anomaly Detection with Domain Shift: A Real-world Dataset and Masked Multi-scale Reconstruction