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

Building video-language foundation models is costly and difficult due to theredundant nature of video data and the lack of high-quality video-languagedatasets. In this paper, we propose an efficient framework to harvest videofoundation models from image ones. Our method is intuitively simple by randomlydropping input video patches and masking out input text during thepost-pretraining procedure. The patch dropping boosts the training efficiencysignificantly and text masking enforces the learning of cross-modal fusion. Weconduct extensive experiments to validate the effectiveness of our method on awide range of video-language downstream tasks including various zero-shottasks, video question answering, and video-text retrieval. Despite itssimplicity, our method achieves state-of-the-art performances, which arecomparable to some heavily pretrained video foundation models. Our method isextremely efficient and can be trained in less than one day on 8 GPUs,requiring only WebVid-10M as pretraining data. We hope our method can serve asa simple yet strong counterpart for prevalent video foundation models, provideuseful insights when building them, and make large pretrained models moreaccessible and sustainable. This is part of the InternVideo project\url{https://github.com/OpenGVLab/InternVideo}.

Harvest Video Foundation Models via Efficient Post-Pretraining