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

As a pioneering work exploring transformer architecture for 3D point cloudunderstanding, Point Transformer achieves impressive results on multiple highlycompetitive benchmarks. In this work, we analyze the limitations of the PointTransformer and propose our powerful and efficient Point Transformer V2 modelwith novel designs that overcome the limitations of previous work. Inparticular, we first propose group vector attention, which is more effectivethan the previous version of vector attention. Inheriting the advantages ofboth learnable weight encoding and multi-head attention, we present a highlyeffective implementation of grouped vector attention with a novel groupedweight encoding layer. We also strengthen the position information forattention by an additional position encoding multiplier. Furthermore, we designnovel and lightweight partition-based pooling methods which enable betterspatial alignment and more efficient sampling. Extensive experiments show thatour model achieves better performance than its predecessor and achievesstate-of-the-art on several challenging 3D point cloud understandingbenchmarks, including 3D point cloud segmentation on ScanNet v2 and S3DIS and3D point cloud classification on ModelNet40. Our code will be available athttps://github.com/Gofinge/PointTransformerV2.

Point Transformer V2: Grouped Vector Attention and Partition-based Pooling