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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Multimedia information have strong temporal correlations that shape the waymodalities co-occur over time. In this paper we study the dynamic nature ofmultimedia and social-media information, where the temporal dimension emergesas a strong source of evidence for learning the temporal correlations acrossvisual and textual modalities. So far, cross-media retrieval models, exploredthe correlations between different modalities (e.g. text and image) to learn acommon subspace, in which semantically similar instances lie in the sameneighbourhood. Building on such knowledge, we propose a novel temporalcross-media neural architecture, that departs from standard cross-mediamethods, by explicitly accounting for the temporal dimension through temporalsubspace learning. The model is softly-constrained with temporal andinter-modality constraints that guide the new subspace learning task byfavouring temporal correlations between semantically similar and temporallyclose instances. Experiments on three distinct datasets show that accountingfor time turns out to be important for cross-media retrieval. Namely, theproposed method outperforms a set of baselines on the task of temporalcross-media retrieval, demonstrating its effectiveness for performing temporalsubspace learning.

Temporal Cross-Media Retrieval with Soft-Smoothing