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

Training supervised video captioning model requires coupled video-captionpairs. However, for many targeted languages, sufficient paired data are notavailable. To this end, we introduce the unpaired video captioning task aimingto train models without coupled video-caption pairs in target language. Tosolve the task, a natural choice is to employ a two-step pipeline system: firstutilizing video-to-pivot captioning model to generate captions in pivotlanguage and then utilizing pivot-to-target translation model to translate thepivot captions to the target language. However, in such a pipeline system, 1)visual information cannot reach the translation model, generating visualirrelevant target captions; 2) the errors in the generated pivot captions willbe propagated to the translation model, resulting in disfluent target captions.To address these problems, we propose the Unpaired Video Captioning with VisualInjection system (UVC-VI). UVC-VI first introduces the Visual Injection Module(VIM), which aligns source visual and target language domains to inject thesource visual information into the target language domain. Meanwhile, VIMdirectly connects the encoder of the video-to-pivot model and the decoder ofthe pivot-to-target model, allowing end-to-end inference by completely skippingthe generation of pivot captions. To enhance the cross-modality injection ofthe VIM, UVC-VI further introduces a pluggable video encoder, i.e., MultimodalCollaborative Encoder (MCE). The experiments show that UVC-VI outperformspipeline systems and exceeds several supervised systems. Furthermore, equippingexisting supervised systems with our MCE can achieve 4% and 7% relative marginson the CIDEr scores to current state-of-the-art models on the benchmark MSVDand MSR-VTT datasets, respectively.

Aligning Source Visual and Target Language Domains for Unpaired Video Captioning