Handling Object Symmetries in CNN-based Pose Estimation
This work is significant for researchers and practitioners in computer vision working on object pose estimation, particularly for improving the accuracy and robustness of CNNs when dealing with symmetric objects, which is an incremental improvement.
This paper addresses the challenges CNN-based pose estimators face with symmetric objects, identifying that common output representations fail to form a closed loop after each symmetry step, leading to issues with gradient-based optimization. The authors propose a "closed symmetry loop" (csl) representation, which generalizes to 6-DOF and achieves state-of-the-art performance for unrefined RGB-based methods on the T-LESS dataset.
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.