![](https://img-blog.csdnimg.cn/20201014180756913.png?x-oss-process=image/resize,m_fixed,h_64,w_64)
3D点云
文章平均质量分 77
北岭狼人
一个超级疯狂的Pythoner
展开
-
点云补全综述 Comprehensive Review of Deep Learning-Based 3D Point Clouds Completion Processing and Analys
点云补全是由部分点云衍生而来的一个生成和估计问题,在三维计算机视觉的应用中起着至关重要的作用。深度学习(DL)的发展令人印象深刻地提高了点云完成的能力和鲁棒性。但是,已完成的点云的质量仍需进一步提高,以满足实际利用。因此,本工作旨在对各种方法进行全面的调查,包括基于点、基于卷积、基于图和基于生成模型的方法等。本调查总结了这些方法之间的比较,以激发进一步的研究见解。此外,本文综述了常用的数据集,并说明了点云补全的应用。最后,我们还讨论了在这一迅速扩大的领域中可能出现的研究趋势。原创 2022-08-22 16:05:11 · 14062 阅读 · 4 评论 -
论文阅读【CVPR-2022】3D Common Corruptions and Data Augmentation
3D Common Corruptions and Data Augmentation3维通用Corruptions和数据扩增原创 2022-03-29 10:38:08 · 2160 阅读 · 0 评论 -
Variational Relational Point Completion Network 阅读笔记
Variational Relational Point Completion NetworkAbstract 摘要Real-scanned point clouds are often incomplete due to viewpoint, occlusion, and noise. Existing point cloud completion methods tend to generate global shape skeletons and hence lack fine local det原创 2021-08-07 17:36:10 · 931 阅读 · 0 评论 -
2021-08-03
TopNet: Structural Point Cloud Decoder@TOCTopNet: Structural Point Cloud Decoder 文章解读TopNet: Structural Point Cloud DecoderAbstract3D point cloud generation is of great use for 3D scene modeling and understanding. Real-world 3D object point clouds can原创 2021-08-03 15:30:50 · 194 阅读 · 0 评论 -
Cycle4Completion: Unpaired Point Cloud Completion using Cycle Transformation with Missing Region Cod
文章解读 Cycle4Completion: Unpaired Point Cloud Completion using Cycle Transformation with Missing Region Coding摘要(abstract)In this paper, we present a novel unpaired point cloud completion network, named Cycle4Completion, to infer the complete geometries fr原创 2021-08-02 17:49:43 · 303 阅读 · 1 评论