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

We propose a novel memory-based tracker via part-level dense memory andvoting-based retrieval, called DMV. Since deep learning techniques have beenintroduced to the tracking field, Siamese trackers have attracted manyresearchers due to the balance between speed and accuracy. However, most ofthem are based on a single template matching, which limits the performance asit restricts the accessible in-formation to the initial target features. Inthis paper, we relieve this limitation by maintaining an external memory thatsaves the tracking record. Part-level retrieval from the memory also liberatesthe information from the template and allows our tracker to better handle thechallenges such as appearance changes and occlusions. By updating the memoryduring tracking, the representative power for the target object can be enhancedwithout online learning. We also propose a novel voting mechanism for thememory reading to filter out unreliable information in the memory. Wecomprehensively evaluate our tracker on OTB-100,TrackingNet, GOT-10k, LaSOT,and UAV123, which show that our method yields comparable results to thestate-of-the-art methods.

DMV: Visual Object Tracking via Part-level Dense Memory and Voting-based Retrieval