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

Rejecting correspondence outliers enables to boost the correspondencequality, which is a critical step in achieving high point cloud registrationaccuracy. The current state-of-the-art correspondence outlier rejection methodsonly utilize the structure features of the correspondences. However, textureinformation is critical to reject the correspondence outliers in our humanvision system. In this paper, we propose General Multimodal Fusion (GMF) tolearn to reject the correspondence outliers by leveraging both the structureand texture information. Specifically, two cross-attention-based fusion layersare proposed to fuse the texture information from paired images and structureinformation from point correspondences. Moreover, we propose a convolutionalposition encoding layer to enhance the difference between Tokens and enable theencoding feature pay attention to neighbor information. Our position encodinglayer will make the cross-attention operation integrate both local and globalinformation. Experiments on multiple datasets(3DMatch, 3DLoMatch, KITTI) andrecent state-of-the-art models (3DRegNet, DGR, PointDSC) prove that our GMFachieves wide generalization ability and consistently improves the point cloudregistration accuracy. Furthermore, several ablation studies demonstrate therobustness of the proposed GMF on different loss functions, lighting conditionsand noises.The code is available at https://github.com/XiaoshuiHuang/GMF.

GMF: General Multimodal Fusion Framework for Correspondence Outlier Rejection