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

Learning to generate 3D point clouds without 3D supervision is an importantbut challenging problem. Current solutions leverage various differentiablerenderers to project the generated 3D point clouds onto a 2D image plane, andtrain deep neural networks using the per-pixel difference with 2D ground truthimages. However, these solutions are still struggling to fully recover finestructures of 3D shapes, such as thin tubes or planes. To resolve this issue,we propose an unsupervised approach for 3D point cloud generation with finestructures. Specifically, we cast 3D point cloud learning as a 2D projectionmatching problem. Rather than using entire 2D silhouette images as a regularpixel supervision, we introduce structure adaptive sampling to randomly sample2D points within the silhouettes as an irregular point supervision, whichalleviates the consistency issue of sampling from different view angles. Ourmethod pushes the neural network to generate a 3D point cloud whose 2Dprojections match the irregular point supervision from different view angles.Our 2D projection matching approach enables the neural network to learn moreaccurate structure information than using the per-pixel difference, especiallyfor fine and thin 3D structures. Our method can recover fine 3D structures from2D silhouette images at different resolutions, and is robust to differentsampling methods and point number in irregular point supervision. Our methodoutperforms others under widely used benchmarks. Our code, data and models areavailable at https://github.com/chenchao15/2D\_projection\_matching.

Unsupervised Learning of Fine Structure Generation for 3D Point Clouds by 2D Projection Matching