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

Many recently developed object detectors focused on coarse-to-fine frameworkwhich contains several stages that classify and regress proposals fromcoarse-grain to fine-grain, and obtains more accurate detection gradually.Multi-resolution models such as Feature Pyramid Network(FPN) integrateinformation of different levels of resolution and effectively improve theperformance. Previous researches also have revealed that localization can befurther improved by: 1) using fine-grained information which is moretranslational variant; 2) refining local areas which is more focused on localboundary information. Based on these principles, we designed a novel boundaryrefinement architecture to improve localization accuracy by combiningcoarse-to-fine framework with feature pyramid structure, named as PyramidalBounding Box Refinement network(PBRnet), which parameterizes gradually focusedboundary areas of objects and leverages lower-level feature maps to extractfiner local information when refining the predicted bounding boxes. Extensiveexperiments are performed on the MS-COCO dataset. The PBRnet brings asignificant performance gains by roughly 3 point of $mAP$ when added to FPN orLibra R-CNN. Moreover, by treating Cascade R-CNN as a coarse-to-fine detectorand replacing its localization branch by the regressor of PBRnet, it leads anextra performance improvement by 1.5 $mAP$, yielding a total performanceboosting by as high as 5 point of $mAP$.

PBRnet: Pyramidal Bounding Box Refinement to Improve Object Localization Accuracy