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

We aim to address the problem of Natural Language Video Localization(NLVL)-localizing the video segment corresponding to a natural languagedescription in a long and untrimmed video. State-of-the-art NLVL methods arealmost in one-stage fashion, which can be typically grouped into twocategories: 1) anchor-based approach: it first pre-defines a series of videosegment candidates (e.g., by sliding window), and then does classification foreach candidate; 2) anchor-free approach: it directly predicts the probabilitiesfor each video frame as a boundary or intermediate frame inside the positivesegment. However, both kinds of one-stage approaches have inherent drawbacks:the anchor-based approach is susceptible to the heuristic rules, furtherlimiting the capability of handling videos with variant length. While theanchor-free approach fails to exploit the segment-level interaction thusachieving inferior results. In this paper, we propose a novel Boundary ProposalNetwork (BPNet), a universal two-stage framework that gets rid of the issuesmentioned above. Specifically, in the first stage, BPNet utilizes ananchor-free model to generate a group of high-quality candidate video segmentswith their boundaries. In the second stage, a visual-language fusion layer isproposed to jointly model the multi-modal interaction between the candidate andthe language query, followed by a matching score rating layer that outputs thealignment score for each candidate. We evaluate our BPNet on three challengingNLVL benchmarks (i.e., Charades-STA, TACoS and ActivityNet-Captions). Extensiveexperiments and ablative studies on these datasets demonstrate that the BPNetoutperforms the state-of-the-art methods.

Boundary Proposal Network for Two-Stage Natural Language Video Localization