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

This paper considers zero-shot Anomaly Detection (AD), a valuable yetunder-studied task, which performs AD without any reference images of the testobjects. Specifically, we employ a language-guided strategy and propose asimple-yet-effective architecture CLIP-AD, leveraging the superior zero-shotclassification capabilities of the large vision-language model CLIP. A naturalidea for anomaly segmentation is to directly calculate the similarity betweentext/image features, but we observe opposite predictions and irrelevanthighlights in the results. Inspired by the phenomena, we introduce a StagedDual-Path model (SDP) that effectively uses features from various levels andapplies architecture and feature surgery to address these issues. Furthermore,delving beyond surface phenomena, we identify the problem arising frommisalignment of text/image features in the joint embedding space. Thus, weintroduce a fine-tuning strategy by adding linear layers and construct anextended model SDP+, further enhancing the performance. Abundant experimentsdemonstrate the effectiveness of our approach, e.g., on VisA, SDP outperformsSOTA by +1.0/+1.2 in classification/segmentation F1 scores, while SDP+ achieves+1.9/+11.7 improvements.

CLIP-AD: A Language-Guided Staged Dual-Path Model for Zero-shot Anomaly Detection