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

Temporal action localization (TAL) is a prevailing task due to its greatapplication potential. Existing works in this field mainly suffer from twoweaknesses: (1) They often neglect the multi-label case and only focus ontemporal modeling. (2) They ignore the semantic information in class labels andonly use the visual information. To solve these problems, we propose a novelCo-Occurrence Relation Module (CORM) that explicitly models the co-occurrencerelationship between actions. Besides the visual information, it furtherutilizes the semantic embeddings of class labels to model the co-occurrencerelationship. The CORM works in a plug-and-play manner and can be easilyincorporated with the existing sequence models. By considering both visual andsemantic co-occurrence, our method achieves high multi-label relationshipmodeling capacity. Meanwhile, existing datasets in TAL always focus onlow-semantic atomic actions. Thus we construct a challenging multi-labeldataset UCF-Crime-TAL that focuses on high-semantic actions by annotating theUCF-Crime dataset at frame level and considering the semantic overlap ofdifferent events. Extensive experiments on two commonly used TAL datasets,\textit{i.e.}, MultiTHUMOS and TSU, and our newly proposed UCF-Crime-TALdemenstrate the effectiveness of the proposed CORM, which achievesstate-of-the-art performance on these datasets.

Co-Occurrence Matters: Learning Action Relation for Temporal Action Localization