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

Grounding temporal video segments described in natural language querieseffectively and efficiently is a crucial capability needed invision-and-language fields. In this paper, we deal with the fast video temporalgrounding (FVTG) task, aiming at localizing the target segment with high speedand favorable accuracy. Most existing approaches adopt elaborately designedcross-modal interaction modules to improve the grounding performance, whichsuffer from the test-time bottleneck. Although several common space-basedmethods enjoy the high-speed merit during inference, they can hardly capturethe comprehensive and explicit relations between visual and textual modalities.In this paper, to tackle the dilemma of speed-accuracy tradeoff, we propose acommonsense-aware cross-modal alignment (CCA) framework, which incorporatescommonsense-guided visual and text representations into a complementary commonspace for fast video temporal grounding. Specifically, the commonsense conceptsare explored and exploited by extracting the structural semantic informationfrom a language corpus. Then, a commonsense-aware interaction module isdesigned to obtain bridged visual and text features by utilizing the learnedcommonsense concepts. Finally, to maintain the original semantic information oftextual queries, a cross-modal complementary common space is optimized toobtain matching scores for performing FVTG. Extensive results on twochallenging benchmarks show that our CCA method performs favorably againststate-of-the-arts while running at high speed. Our code is available athttps://github.com/ZiyueWu59/CCA.

Learning Commonsense-aware Moment-Text Alignment for Fast Video Temporal Grounding