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

Fully supervised change detection methods have achieved significantadvancements in performance, yet they depend severely on acquiring costlypixel-level labels. Considering that the patch-level annotations also containabundant information corresponding to both changed and unchanged objects inbi-temporal images, an intuitive solution is to segment the changes withpatch-level annotations. How to capture the semantic variations associated withthe changed and unchanged regions from the patch-level annotations to obtainpromising change results is the critical challenge for the weakly supervisedchange detection task. In this paper, we propose a memory-supported transformer(MS-Former), a novel framework consisting of a bi-directional attention block(BAB) and a patch-level supervision scheme (PSS) tailored for weakly supervisedchange detection with patch-level annotations. More specifically, the BAMcaptures contexts associated with the changed and unchanged regions from thetemporal difference features to construct informative prototypes stored in thememory bank. On the other hand, the BAM extracts useful information from theprototypes as supplementary contexts to enhance the temporal differencefeatures, thereby better distinguishing changed and unchanged regions. Afterthat, the PSS guides the network learning valuable knowledge from thepatch-level annotations, thus further elevating the performance. Experimentalresults on three benchmark datasets demonstrate the effectiveness of ourproposed method in the change detection task. The demo code for our work willbe publicly available at \url{https://github.com/guanyuezhen/MS-Former}.

MS-Former: Memory-Supported Transformer for Weakly Supervised Change Detection with Patch-Level Annotations