arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2012.11002cs.CV
收藏

Deep Bingham Networks:处理姿态估计中的不确定性和模糊性

Deep Bingham Networks: Dealing with Uncertainty and Ambiguity in Pose Estimation

Technische Universität München(慕尼黑工业大学);Siemens AG(西门子公司);Stanford University(斯坦福大学)

Haowen Deng, Mai Bui, Nassir Navab, Leonidas Guibas, Slobodan Ilic, Tolga Birdal

摘要:

在这项工作中,我们介绍了深度宾汉网络(DBN),一个通用的框架,可以自然地处理姿态相关的不确定性和模糊性,在几乎所有的现实生活中的应用程序有关的3D数据。虽然现有的作品努力找到一个单一的解决方案的姿态估计问题,我们和平与造成高度不确定性的解决方案,以确定为最好的模糊性。相反,我们报告了一个家庭的姿态,捕捉的解决方案空间的性质。DBN通过(i)可以产生不同分布模式的多假设预测头;以及(ii)受益于旋转上的宾汉分布的新颖损失函数来扩展现有技术的直接姿态回归网络。这样,DBN既可以在提供不确定性信息的明确情况下工作,也可以在需要每个模式的不确定性的模糊场景中工作。在技术方面,我们的网络回归连续Bingham混合模型,适用于图像等2D数据和点云等3D数据。我们提出了新的训练策略,以避免在训练过程中的模式或后崩溃,并提高数值稳定性。我们的方法在两种不同的应用程序上进行了彻底的测试,这两种应用程序采用两种不同的模式:(i)从图像中进行6D相机重新定位;(ii)从3D点云中进行对象姿态估计,表现出比现有技术更好的优势。对于前者,我们贡献了我们自己的数据集,由五个室内场景组成,其中不可避免地要捕获与难以唯一识别的视图对应的图像。对于后者,我们实现了最佳结果,特别是对于ModelNet数据集的对称对象。

英文摘要:

In this work, we introduce Deep Bingham Networks (DBN), a generic framework that can naturally handle pose-related uncertainties and ambiguities arising in almost all real life applications concerning 3D data. While existing works strive to find a single solution to the pose estimation problem, we make peace with the ambiguities causing high uncertainty around which solutions to identify as the best. Instead, we report a family of poses which capture the nature of the solution space. DBN extends the state of the art direct pose regression networks by (i) a multi-hypotheses prediction head which can yield different distribution modes; and (ii) novel loss functions that benefit from Bingham distributions on rotations. This way, DBN can work both in unambiguous cases providing uncertainty information, and in ambiguous scenes where an uncertainty per mode is desired. On a technical front, our network regresses continuous Bingham mixture models and is applicable to both 2D data such as images and to 3D data such as point clouds. We proposed new training strategies so as to avoid mode or posterior collapse during training and to improve numerical stability. Our methods are thoroughly tested on two different applications exploiting two different modalities: (i) 6D camera relocalization from images; and (ii) object pose estimation from 3D point clouds, demonstrating decent advantages over the state of the art. For the former we contributed our own dataset composed of five indoor scenes where it is unavoidable to capture images corresponding to views that are hard to uniquely identify. For the latter we achieve the top results especially for symmetric objects of ModelNet dataset.

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。

↑