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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Robust point cloud registration in real-time is an important prerequisite formany mapping and localization algorithms. Traditional methods like ICP tend tofail without good initialization, insufficient overlap or in the presence ofdynamic objects. Modern deep learning based registration approaches presentmuch better results, but suffer from a heavy run-time. We overcome thesedrawbacks by introducing StickyPillars, a fast, accurate and extremely robustdeep middle-end 3D feature matching method on point clouds. It uses graphneural networks and performs context aggregation on sparse 3D key-points withthe aid of transformer based multi-head self and cross-attention. The networkoutput is used as the cost for an optimal transport problem whose solutionyields the final matching probabilities. The system does not rely on handcrafted feature descriptors or heuristic matching strategies. We presentstate-of-art art accuracy results on the registration problem demonstrated onthe KITTI dataset while being four times faster then leading deep methods.Furthermore, we integrate our matching system into a LiDAR odometry pipelineyielding most accurate results on the KITTI odometry dataset. Finally, wedemonstrate robustness on KITTI odometry. Our method remains stable in accuracywhere state-of-the-art procedures fail on frame drops and higher speeds.

StickyPillars: Robust and Efficient Feature Matching on Point Clouds using Graph Neural Networks