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

Semantic understanding of 3D point cloud relies on learning models withmassively annotated data, which, in many cases, are expensive or difficult tocollect. This has led to an emerging research interest in semi-supervisedlearning (SSL) for 3D point cloud. It is commonly assumed in SSL that theunlabeled data are drawn from the same distribution as that of the labeledones; This assumption, however, rarely holds true in realistic environments.Blindly using out-of-distribution (OOD) unlabeled data could harm SSLperformance. In this work, we propose to selectively utilize unlabeled datathrough sample weighting, so that only conducive unlabeled data would beprioritized. To estimate the weights, we adopt a bi-level optimizationframework which iteratively optimizes a metaobjective on a held-out validationset and a task-objective on a training set. Faced with the instability ofefficient bi-level optimizers, we further propose three regularizationtechniques to enhance the training stability. Extensive experiments on 3D pointcloud classification and segmentation tasks verify the effectiveness of ourproposed method. We also demonstrate the feasibility of a more efficienttraining strategy.

Open-Set Semi-Supervised Learning for 3D Point Cloud Understanding