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

Temporal sentence grounding (TSG) aims to locate a specific moment from anuntrimmed video with a given natural language query. Recently, weaklysupervised methods still have a large performance gap compared to fullysupervised ones, while the latter requires laborious timestamp annotations. Inthis study, we aim to reduce the annotation cost yet keep competitiveperformance for TSG task compared to fully supervised ones. To achieve thisgoal, we investigate a recently proposed glance-supervised temporal sentencegrounding task, which requires only single frame annotation (referred to asglance annotation) for each query. Under this setup, we propose a DynamicGaussian prior based Grounding framework with Glance annotation (D3G), whichconsists of a Semantic Alignment Group Contrastive Learning module (SA-GCL) anda Dynamic Gaussian prior Adjustment module (DGA). Specifically, SA-GCL samplesreliable positive moments from a 2D temporal map via jointly leveragingGaussian prior and semantic consistency, which contributes to aligning thepositive sentence-moment pairs in the joint embedding space. Moreover, toalleviate the annotation bias resulting from glance annotation and modelcomplex queries consisting of multiple events, we propose the DGA module, whichadjusts the distribution dynamically to approximate the ground truth of targetmoments. Extensive experiments on three challenging benchmarks verify theeffectiveness of the proposed D3G. It outperforms the state-of-the-art weaklysupervised methods by a large margin and narrows the performance gap comparedto fully supervised methods. Code is available athttps://github.com/solicucu/D3G.

D3G: Exploring Gaussian Prior for Temporal Sentence Grounding with Glance Annotation