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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

This paper proposes a new transformer-based framework to learn class-specificobject localization maps as pseudo labels for weakly supervised semanticsegmentation (WSSS). Inspired by the fact that the attended regions of theone-class token in the standard vision transformer can be leveraged to form aclass-agnostic localization map, we investigate if the transformer model canalso effectively capture class-specific attention for more discriminativeobject localization by learning multiple class tokens within the transformer.To this end, we propose a Multi-class Token Transformer, termed as MCTformer,which uses multiple class tokens to learn interactions between the class tokensand the patch tokens. The proposed MCTformer can successfully produceclass-discriminative object localization maps from class-to-patch attentionscorresponding to different class tokens. We also propose to use a patch-levelpairwise affinity, which is extracted from the patch-to-patch transformerattention, to further refine the localization maps. Moreover, the proposedframework is shown to fully complement the Class Activation Mapping (CAM)method, leading to remarkably superior WSSS results on the PASCAL VOC and MSCOCO datasets. These results underline the importance of the class token forWSSS.

Multi-class Token Transformer for Weakly Supervised Semantic Segmentation