arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1909.02533cs.CVcs.GR
收藏

C3DPO:非刚性运动恢复结构的规范三维姿态网络

C3DPO: Canonical 3D Pose Networks for Non-Rigid Structure From Motion

Facebook AI Research(Facebook AI研究院)

David Novotny, Nikhila Ravi, Benjamin Graham, Natalia Neverova, Andrea Vedaldi

摘要:

我们提出了C3DPO,一种从无约束图像中的2D关键点注释中提取可变形对象的3D模型的方法。我们通过学习一个深度网络来实现这一点,该网络一次从单个视图重建3D对象,考虑部分遮挡,并明确考虑视点变化和对象变形的影响。为了实现这种因式分解,我们引入了一种新的正则化技术。我们首先表明,因子分解是成功的,当且仅当,存在一定的正则化函数的重建形状。然后,我们一起学习规范化函数和重构函数,这限制了结果的一致性。我们展示了最先进的重建结果的方法,不使用地面实况3D监督的一些基准,包括Up3D和PASCAL3D+。源代码已在https://github.com/facebookresearch/c3dpo_nrsfm上提供。

英文摘要:

We propose C3DPO, a method for extracting 3D models of deformable objects from 2D keypoint annotations in unconstrained images. We do so by learning a deep network that reconstructs a 3D object from a single view at a time, accounting for partial occlusions, and explicitly factoring the effects of viewpoint changes and object deformations. In order to achieve this factorization, we introduce a novel regularization technique. We first show that the factorization is successful if, and only if, there exists a certain canonicalization function of the reconstructed shapes. Then, we learn the canonicalization function together with the reconstruction one, which constrains the result to be consistent. We demonstrate state-of-the-art reconstruction results for methods that do not use ground-truth 3D supervision for a number of benchmarks, including Up3D and PASCAL3D+. Source code has been made available at https://github.com/facebookresearch/c3dpo_nrsfm.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑