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 post-processor, called NeighborTrack, that leverages neighborinformation of the tracking target to validate and improve single-objecttracking (SOT) results. It requires no additional data or retraining. Instead,it uses the confidence score predicted by the backbone SOT network toautomatically derive neighbor information and then uses this information toimprove the tracking results. When tracking an occluded target, its appearancefeatures are untrustworthy. However, a general siamese network often cannottell whether the tracked object is occluded by reading the confidence scorealone, because it could be misled by neighbors with high confidence scores. Ourproposed NeighborTrack takes advantage of unoccluded neighbors' information toreconfirm the tracking target and reduces false tracking when the target isoccluded. It not only reduces the impact caused by occlusion, but also fixestracking problems caused by object appearance changes. NeighborTrack isagnostic to SOT networks and post-processing methods. For the VOT challengedataset commonly used in short-term object tracking, we improve three famousSOT networks, Ocean, TransT, and OSTrack, by an average of ${1.92\%}$ EAO and${2.11\%}$ robustness. For the mid- and long-term tracking experiments based onOSTrack, we achieve state-of-the-art ${72.25\%}$ AUC on LaSOT and ${75.7\%}$ AOon GOT-10K. Code duplication can be found inhttps://github.com/franktpmvu/NeighborTrack.

NeighborTrack: Improving Single Object Tracking by Bipartite Matching with Neighbor Tracklets