Revisiting the Continuity of Rotation Representations in Neural Networks

In this paper, we provide some careful analysis of certain pathological behavior of Euler angles and unit quaternions encountered in previous works related to rotation representation in neural networks. In particular, we show that for certain problems, these two representations will provably produce completely wrong results for some inputs, and that this behavior is inherent in the topological property of the problem itself and is not caused by unsuitable network architectures or training procedures. We further show that previously proposed embeddings of $\mathrm{SO}(3)$ into higher dimensional Euclidean spaces aimed at fixing this behavior are not universally effective, due to possible symmetry in the input causing changes to the topology of the input space. We propose an ensemble trick as an alternative solution.

[1]  Ruben Villegas,et al.  Neural Kinematic Networks for Unsupervised Motion Retargetting , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[2]  Yi Zhou,et al.  On the Continuity of Rotation Representations in Neural Networks , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  R. Ho.  Algebraic Topology , 2022 .

[4]  Leonidas J. Guibas,et al.  PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[5]  Ashutosh Saxena,et al.  Learning 3-D object orientation from images , 2009, 2009 IEEE International Conference on Robotics and Automation.

[6]  Floris Takens,et al.  The minimal number of critical points of a function on a compact manifold and the Lusternik-Schnirelman category , 1968 .

[7]  J. Mebius.  A MATRIX-BASED PROOF OF THE QUATERNION REPRESENTATION THEOREM FOR FOUR-DIMENSIONAL ROTATIONS , 2005, math/0501249.

[8]  Dieter Fox,et al.  PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes , 2017, Robotics: Science and Systems.

[9]  R. E. Miles.  On random rotations in R3 , 1965 .

[10]  Robert H. Halstead,et al.  Matrix Computations , 2011, Encyclopedia of Parallel Computing.