Learning Multi-View Camera Relocalization With Graph Neural Networks
Authors:
- 1 June 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 11372-11381
Abstract
We propose to construct a view graph to excavate the information of the whole given sequence for absolute camera pose estimation. Specifically, we harness GNNs to model the graph, allowing even non-consecutive frames to exchange information with each other. Rather than adopting the regular GNNs directly, we redefine the nodes, edges, and embedded functions to fit the relocalization task. Redesigned GNNs cooperate with CNNs in guiding knowledge propagation and feature extraction respectively to process multi-view high-dimension image features iteratively at different levels. Besides, a general graph-based loss function beyond constraints between consecutive views is employed for training the network in an end-to-end fashion. Extensive experiments conducted on both indoor and outdoor datasets demonstrate that our method outperforms previous approaches especially in large-scale and challenging scenarios.Keywords
This publication has 32 references indexed in Scilit:
- Efficient & Effective Prioritized Matching for Large-Scale Image-Based LocalizationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016
- Large-Scale Location Recognition and the Geometric Burstiness Problem2016
- Deep Residual Learning for Image Recognition2016
- Structure-from-Motion Revisited2016
- NetVLAD: CNN Architecture for Weakly Supervised Place Recognition2016
- Modelling uncertainty in deep learning for camera relocalization2016
- Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification2015
- PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization2015
- 24/7 place recognition by view synthesis2015
- Scene Coordinate Regression Forests for Camera Relocalization in RGB-D Images2013