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

While point-based neural architectures have demonstrated their efficacy, thetime-consuming sampler currently prevents them from performing real-timereasoning on scene-level point clouds. Existing methods attempt to overcomethis issue by using random sampling strategy instead of the commonly-adoptedfarthest point sampling~(FPS), but at the expense of lower performance. So theeffectiveness/efficiency trade-off remains under-explored. In this paper, wereveal the key to high-quality sampling is ensuring an even spacing betweenpoints in the subset, which can be naturally obtained through a grid. Based onthis insight, we propose a hierarchical adaptive voxel-guided point samplerwith linear complexity and high parallelization for real-time applications.Extensive experiments on large-scale point cloud detection and segmentationtasks demonstrate that our method achieves competitive performance with themost powerful FPS, at an amazing speed that is more than 100 times faster. Thisbreakthrough in efficiency addresses the bottleneck of the sampling step whenhandling scene-level point clouds. Furthermore, our sampler can be easilyintegrated into existing models and achieves a 20$\sim$80\% reduction inruntime with minimal effort. The code will be available athttps://github.com/OuyangJunyuan/pointcloud-3d-detector-tensorrt

Hierarchical Adaptive Voxel-guided Sampling for Real-time Applications in Large-scale Point Clouds