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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Model update lies at the heart of object tracking. Generally, model update isformulated as an online learning problem where a target model is learned overthe online training set. Our key innovation is to \emph{formulate the modelupdate problem in the meta-learning framework and learn the online learningalgorithm itself using large numbers of offline videos}, i.e., \emph{learningto update}. The learned updater takes as input the online training set andoutputs an updated target model. As a first attempt, we design the learnedupdater based on recurrent neural networks (RNNs) and demonstrate itsapplication in a template-based tracker and a correlation filter-based tracker.Our learned updater consistently improves the base trackers and runs fasterthan realtime on GPU while requiring small memory footprint during testing.Experiments on standard benchmarks demonstrate that our learned updateroutperforms commonly used update baselines including the efficient exponentialmoving average (EMA)-based update and the well-designed stochastic gradientdescent (SGD)-based update. Equipped with our learned updater, thetemplate-based tracker achieves state-of-the-art performance among realtimetrackers on GPU.

Learning to Update for Object Tracking with Recurrent Meta-learner