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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Building a universal Video-Language model for solving various videounderstanding tasks (\emph{e.g.}, text-video retrieval, video questionanswering) is an open challenge to the machine learning field. Towards thisgoal, most recent works build the model by stacking uni-modal and cross-modalfeature encoders and train it with pair-wise contrastive pre-text tasks. Thoughoffering attractive generality, the resulted models have to compromise betweenefficiency and performance. They mostly adopt different architectures to dealwith different downstream tasks. We find this is because the pair-wise trainingcannot well \emph{align} and \emph{fuse} features from different modalities. Wethen introduce \textbf{Clover}\textemdash a Correlated Video-Languagepre-training method\textemdash towards a universal Video-Language model forsolving multiple video understanding tasks with neither performance norefficiency compromise. It improves cross-modal feature alignment and fusion viaa novel tri-modal alignment pre-training task. Additionally, we propose toenhance the tri-modal alignment via incorporating learning from semantic maskedsamples and a new pair-wise ranking loss. Clover establishes newstate-of-the-arts on multiple downstream tasks, including three retrieval tasksfor both zero-shot and fine-tuning settings, and eight video question answeringtasks. Codes and pre-trained models will be released at\url{https://github.com/LeeYN-43/Clover}.

Clover: Towards A Unified Video-Language Alignment and Fusion Model