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

Change detection (CD) by comparing two bi-temporal images is a crucial taskin remote sensing. With the advantages of requiring no cumbersome labeledchange information, unsupervised CD has attracted extensive attention in thecommunity. However, existing unsupervised CD approaches rarely consider theseasonal and style differences incurred by the illumination and atmosphericconditions in multi-temporal images. To this end, we propose a change detectionwith domain shift setting for remote sensing images. Furthermore, we present anovel unsupervised CD method using a light-weight transformer, calledUCDFormer. Specifically, a transformer-driven image translation composed of alight-weight transformer and a domain-specific affinity weight is firstproposed to mitigate domain shift between two images with real-time efficiency.After image translation, we can generate the difference map between thetranslated before-event image and the original after-event image. Then, a novelreliable pixel extraction module is proposed to select significantlychanged/unchanged pixel positions by fusing the pseudo change maps of fuzzyc-means clustering and adaptive threshold. Finally, a binary change map isobtained based on these selected pixel pairs and a binary classifier.Experimental results on different unsupervised CD tasks with seasonal and stylechanges demonstrate the effectiveness of the proposed UCDFormer. For example,compared with several other related methods, UCDFormer improves performance onthe Kappa coefficient by more than 12\%. In addition, UCDFormer achievesexcellent performance for earthquake-induced landslide detection whenconsidering large-scale applications. The code is available at\url{https://github.com/zhu-xlab/UCDFormer}

UCDFormer: Unsupervised Change Detection Using a Transformer-driven Image Translation