Papers of the day   All papers

On the Continuity of Rotation Representations in Neural Networks

Comments

Jan 26 2020 Ankur Handa

Quaternions and Euler angles are discontinuous and difficult for neural networks to learn. They show 3D rotations have continuous representations in 5D and 6D, which are more suitable for learning. i.e. regress two vectors and apply Graham-Schmidt (GS). https://arxiv.org/abs/1812.07035 https://t.co/fXUF3sgkTT
11 replies, 667 likes


Jan 26 2020 Matt Miesnieks

So this is literally 6D AI ? :) @6d_ai
8 replies, 69 likes


Jan 27 2020 Marc B. Reynolds

This keeps show up in my timeline, so: Quat's are not discontinuous by any normal definition.
4 replies, 14 likes


Jul 12 2019 Chris Choy

I really like this paper from Hao Li's group: On the Continuity of Rotation Representations in Neural Networks, CVPR19 :) https://arxiv.org/pdf/1812.07035.pdf
0 replies, 12 likes


Content