[2010.04642] Torch-Points3D: A Modular Multi-Task Frameworkfor Reproducible Deep Learning on 3D Point Cloudsopen searchopen navigation menucontact arXivsubscribe to arXiv mailings

We introduce Torch-Points3D, an open-source framework designed to facilitate the use of deep networks on3D data. Its modular design, efficient implementation, and user-friendly interfaces make it a relevant tool for research and productization alike. Beyond multiple quality-of-life features, our goal is to standardize a higher level of transparency and reproducibility in 3D deep learning research, and to lower its barrier to entry. In this paper, we present the design principles of Torch-Points3D, as well as extensive benchmarks of multiple state-of-the-art algorithms and inference schemes across several datasets and tasks. The modularity of Torch-Points3D allows us to design fair and rigorous experimental protocols in which all methods are evaluated in the same conditions. The Torch-Points3D repository :https://github.com/nicolas-chaulet/torch-points3d

Keywords: deep learning
Date: 2020/10/14 06:53

Related Entries

Read more [2001.10773] Virtual KITTI 2contact arXivarXiv Twitter
0 users, 2 mentions 2020/01/31 00:51
Read more [2005.11037] Style Normalization and Restitution for Generalizable Person Re-identificationopen sear...
0 users, 2 mentions 2020/05/27 14:21
Read more [2007.10983] Learning Monocular Visual Odometry via Self-Supervised Long-Term Modelingopen searchope...
0 users, 3 mentions 2020/07/22 15:51
Read more [2009.00564] A Primer on Motion Capture with Deep Learning: Principles, Pitfalls and Perspectivesope...
0 users, 3 mentions 2020/09/05 18:52
Read more [2009.07724] Evaluating Self-Supervised Pretraining Without Using Labelsopen searchopen navigation m...
0 users, 5 mentions 2020/09/18 14:22