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

Towards 3D object tracking in point clouds, a novel point-to-box networktermed P2B is proposed in an end-to-end learning manner. Our main idea is tofirst localize potential target centers in 3D search area embedded with targetinformation. Then point-driven 3D target proposal and verification are executedjointly. In this way, the time-consuming 3D exhaustive search can be avoided.Specifically, we first sample seeds from the point clouds in template andsearch area respectively. Then, we execute permutation-invariant featureaugmentation to embed target clues from template into search area seeds andrepresent them with target-specific features. Consequently, the augmentedsearch area seeds regress the potential target centers via Hough voting. Thecenters are further strengthened with seed-wise targetness scores. Finally,each center clusters its neighbors to leverage the ensemble power for joint 3Dtarget proposal and verification. We apply PointNet++ as our backbone andexperiments on KITTI tracking dataset demonstrate P2B's superiority (~10%'simprovement over state-of-the-art). Note that P2B can run with 40FPS on asingle NVIDIA 1080Ti GPU. Our code and model are available athttps://github.com/HaozheQi/P2B.

P2B: Point-to-Box Network for 3D Object Tracking in Point Clouds