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

Videos are a rich source for self-supervised learning (SSL) of visualrepresentations due to the presence of natural temporal transformations ofobjects. However, current methods typically randomly sample video clips forlearning, which results in an imperfect supervisory signal. In this work, wepropose PreViTS, an SSL framework that utilizes an unsupervised tracking signalfor selecting clips containing the same object, which helps better utilizetemporal transformations of objects. PreViTS further uses the tracking signalto spatially constrain the frame regions to learn from and trains the model tolocate meaningful objects by providing supervision on Grad-CAM attention maps.To evaluate our approach, we train a momentum contrastive (MoCo) encoder onVGG-Sound and Kinetics-400 datasets with PreViTS. Training with PreViTSoutperforms representations learnt by contrastive strategy alone on videodownstream tasks, obtaining state-of-the-art performance on actionclassification. PreViTS helps learn feature representations that are morerobust to changes in background and context, as seen by experiments on datasetswith background changes. Learning from large-scale videos with PreViTS couldlead to more accurate and robust visual feature representations.

PreViTS: Contrastive Pretraining with Video Tracking Supervision