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

Video grounding aims to localize the target moment in an untrimmed videocorresponding to a given sentence query. Existing methods typically select thebest prediction from a set of predefined proposals or directly regress thetarget span in a single-shot manner, resulting in the absence of a systematicalprediction refinement process. In this paper, we propose DiffusionVG, a novelframework with diffusion models that formulates video grounding as aconditional generation task, where the target span is generated from Gaussiannoise inputs and interatively refined in the reverse diffusion process. Duringtraining, DiffusionVG progressively adds noise to the target span with a fixedforward diffusion process and learns to recover the target span in the reversediffusion process. In inference, DiffusionVG can generate the target span fromGaussian noise inputs by the learned reverse diffusion process conditioned onthe video-sentence representations. Without bells and whistles, our DiffusionVGdemonstrates superior performance compared to existing well-crafted models onmainstream Charades-STA, ActivityNet Captions and TACoS benchmarks.

Exploring Iterative Refinement with Diffusion Models for Video Grounding