arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2011.13209cs.CVcs.RO
收藏

基于CNN的姿态估计中的物体对称性处理

Handling Object Symmetries in CNN-based Pose Estimation

Faculty of Mathematics and Computer Science, University of Bremen(不来梅大学数学与计算机科学学院)

Jesse Richter-Klug, Udo Frese

AI总结:

针对基于CNN的姿态估计器处理对称物体时的表征不连续问题,提出闭环对称(csl)表征及对称等价姿态消歧方法,适配连续与离散旋转对称,在T-LESS数据集上达无优化RGB方法的SOTA水平。

摘要:

本文研究了基于卷积神经网络(Convolutional Neural Networks,CNN)的姿态估计器在处理对称物体时面临的问题。我们分析了物体连续旋转时CNN输出表征的取值,发现其在每一步对称操作后必须形成闭环,否则本身作为连续函数的CNN就需要拟合一个不连续函数。通过一个1自由度(1-DOF)的简单示例,我们证明常用表征无法满足这一要求,并分析了由此引发的问题。我们特别发现,常用于构建对称性感知损失的“最小对称误差(min-over-symmetries)”方法,在基于梯度的优化(即深度学习)中效果往往不佳。\n基于这些发现,我们提出了名为“闭环对称(closed symmetry loop,csl)”的表征:将相关向量的角度乘以对称阶数,随后将其推广到6自由度(6-DOF)场景。该表征拓展了我们在[Richter-Klug, ICVS, 2019]中提出的算法,新增了在最终姿态估计阶段区分对称等价姿态的方法,可处理连续旋转对称(如瓶子)和离散旋转对称(如4重对称盒子)两类情况。我们在T-LESS数据集上对算法进行了评估,其性能达到了无后处理优化的基于RGB的方法的当前最优水平。

英文摘要:

In this paper, we investigate the problems that Convolutional Neural Networks (CNN)-based pose estimators have with symmetric objects. We considered the value of the CNN's output representation when continuously rotating the object and found that it has to form a closed loop after each step of symmetry. Otherwise, the CNN (which is itself a continuous function) has to replicate an uncontinuous function. On a 1-DOF toy example we show that commonly used representations do not fulfill this demand and analyze the problems caused thereby. In particular, we find that the popular min-over-symmetries approach for creating a symmetry-aware loss tends not to work well with gradient-based optimization, i.e. deep learning. We propose a representation called "closed symmetry loop" (csl) from these insights, where the angle of relevant vectors is multiplied by the symmetry order and then generalize it to 6-DOF. The representation extends our algorithm from [Richter-Klug, ICVS, 2019] including a method to disambiguate symmetric equivalents during the final pose estimation. The algorithm handles continuous rotational symmetry (e.g. a bottle) and discrete rotational symmetry (e.g. a 4-fold symmetric box). It is evaluated on the T-LESS dataset, where it reaches state-of-the-art for unrefining RGB-based methods.

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。

↑