.png){kind=icon}
返回
Learning 3D Shape Completion Under Weak Supervision
.png){kind=icon}
OA
AI
DOI：10.1007/s11263-018-1126-y.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
We address the problem of 3D shape completion from sparse and noisy point clouds, a fundamental problem in computer vision and robotics. Recent approaches are either data-driven or learning-based: Data-driven approaches rely on a shape model whose parameters are optimized to fit the observations; Learning-based approaches, in contrast, avoid the expensive optimization step by learning to directly predict complete shapes from incomplete observations in a fully-supervised setting. However, full supervision is often not available in practice. In this work, we propose a weakly-supervised learning-based approach to 3D shape completion which neither requires slow optimization nor direct supervision. While we also learn a shape prior on synthetic data, we amortize, i.e., learn, maximum likelihood fitting using deep neural networks resulting in efficient shape completion without sacrificing accuracy. On synthetic benchmarks based on ShapeNet (Chang et al. Shapenet: an information-rich 3d model repository, 2015. ) and ModelNet (Wu et al., in: Proceedings of IEEE conference on computer vision and pattern recognition (CVPR), 2015) as well as on real robotics data from KITTI (Geiger et al., in: Proceedings of IEEE conference on computer vision and pattern recognition (CVPR), 2012) and Kinect (Yang et al., 3d object dense reconstruction from a single depth view, 2018. ), we demonstrate that the proposed amortized maximum likelihood approach is able to compete with the fully supervised baseline of Dai et al. (in: Proceedings of IEEE conference on computer vision and pattern recognition (CVPR), 2017) and outperforms the data-driven approach of Engelmann et al. (in: Proceedings of the German conference on pattern recognition (GCPR), 2016), while requiring less supervision and being significantly faster.
Keyword:
3D shape completion
3D reconstruction
Weakly-supervised learning
Amortized inference
Benchmark
AI总结
对已上传原文的论文进行重点信息的提取，主要内容包括：简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。
期刊
IF:
9.3
论文数:
3.9K
被引数:
2.8W
机构
引用论文
Inducible protein-10 as a predictive marker of antiviral hepatitis C treatment: A systematic review
OAAI
OAAI分享收藏分享收藏分享收藏Impact of material emissions and sorption of volatile organic compounds on indoor air quality in a low energy building: Field measurements and modeling.png)
Indoor AirOAAI
Indoor AirIF0
OAAI分享收藏分享收藏Effect of poly(lactic acid) crystallization on its mechanical and heat resistance performancesPolymer
PolymerIF0
PREAI
分享收藏Analysis of the formation of monodisperse populations by homogeneous nucleation通过均相成核分析单分散种群的形成
PREAI
分享收藏
查看更多内容