How to improve CNN-based 6-DoF camera pose estimation
Abstract
Convolutional neural networks (CNNs) and transfer learning have recently been used for 6 degrees of freedom (6-DoF) camera pose estimation. While they do not reach the same accuracy as visual SLAM-based approaches and are restricted to a specific environment, they excel in robustness and can be applied even to a single image. In this paper, we study PoseNet [1] and investigate modifications based on datasets' characteristics to improve the accuracy of the pose estimates. In particular, we emphasize the importance of field-of-view over image resolution; we present a data augmentation scheme to reduce overfitting; we study the effect of Long-Short-Term-Memory (LSTM) cells. Lastly, we combine these modifications and improve PoseNet's performance for monocular CNN based camera pose regression.
Cite
@article{arxiv.1909.10312,
title = {How to improve CNN-based 6-DoF camera pose estimation},
author = {Soroush Seifi and Tinne Tuytelaars},
journal= {arXiv preprint arXiv:1909.10312},
year = {2019}
}
Comments
Accepted at Deep Learning for Visual SLAM workshop at ICCV 2019