On Representation of 3D Rotation in the Context of Deep Learning

TL;DR

This paper compares different 3D rotation representations in deep learning, showing that continuous 5D and 6D representations outperform others in rotation estimation tasks across synthetic and real datasets.

Contribution

It provides a comprehensive evaluation of rotation representations and their effects on neural network performance in 3D rotation estimation.

Findings

5D and 6D representations outperform discontinuous ones

Continuous representations improve learning stability

Texture and data distribution influence estimation accuracy

Abstract

This paper investigates various methods of representing 3D rotations and their impact on the learning process of deep neural networks. We evaluated the performance of ResNet18 networks for 3D rotation estimation using several rotation representations and loss functions on both synthetic and real data. The real datasets contained 3D scans of industrial bins, while the synthetic datasets included views of a simple asymmetric object rendered under different rotations. On synthetic data, we also assessed the effects of different rotation distributions within the training and test sets, as well as the impact of the object's texture. In line with previous research, we found that networks using the continuous 5D and 6D representations performed better than the discontinuous ones.