How to improve CNN-based 6-DoF camera pose estimation

2019/09/23 by Seifi, Soroush, Tuytelaars, Tinne
#Artificial Intelligence (cs.AI) #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Robotics (cs.RO)

paper · doi:10.48550/arxiv.1909.10312

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.

Related

• PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization
• FAR: Flexible, Accurate and Robust 6DoF Relative Camera Pose Estimation
• Understanding the Limitations of CNN-based Absolute Camera Pose\n Regression
• Reassessing the Limitations of CNN Methods for Camera Pose Regression
• Geometric Loss Functions for Camera Pose Regression with Deep Learning