Indoor Relocalization in Challenging Environments With Dual-Stream Convolutional Neural Networks

@article{Li2018IndoorRI,
  title={Indoor Relocalization in Challenging Environments With Dual-Stream Convolutional Neural Networks},
  author={Ruihao Li and Qiang Liu and Jianjun Gui and Dongbing Gu and Huosheng Hu},
  journal={IEEE Transactions on Automation Science and Engineering},
  year={2018},
  volume={15},
  pages={651-662}
}
This paper presents an indoor relocalization system using a dual-stream convolutional neural network (CNN) with both color images and depth images as the network inputs. Aiming at the pose regression problem, a deep neural network architecture for RGB-D images is introduced, a training method by stages for the dual-stream CNN is presented, different depth image encoding methods are discussed, and a novel encoding method is proposed. By introducing the range information into the network through… CONTINUE READING

Citations

Publications citing this paper.
SHOWING 1-10 OF 27 CITATIONS

Probabilistic Projective Association and Semantic Guided Relocalization for Dense Reconstruction

  • 2019 International Conference on Robotics and Automation (ICRA)
  • 2019
VIEW 5 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

End-to-end Network for Monocular Visual Odometry Based on Image Sequence

  • 2019 International Joint Conference on Neural Networks (IJCNN)
  • 2019
VIEW 2 EXCERPTS
CITES BACKGROUND

Long-Term Knowledge Distillation of Visual Place Classifiers

  • 2019 IEEE Intelligent Transportation Systems Conference (ITSC)
  • 2019
VIEW 1 EXCERPT
CITES METHODS

References

Publications referenced by this paper.
SHOWING 1-10 OF 35 REFERENCES

Modelling uncertainty in deep learning for camera relocalization

  • 2016 IEEE International Conference on Robotics and Automation (ICRA)
  • 2015
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization

  • 2015 IEEE International Conference on Computer Vision (ICCV)
  • 2015
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

Method for registration of 3-D shapes

  • Other Conferences
  • 1992
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

ORB-SLAM: A Versatile and Accurate Monocular SLAM System

  • IEEE Transactions on Robotics
  • 2015
VIEW 9 EXCERPTS
HIGHLY INFLUENTIAL

Appearance-only SLAM at large scale with FAB-MAP 2.0

  • I. J. Robotics Res.
  • 2011
VIEW 8 EXCERPTS
HIGHLY INFLUENTIAL

Simultaneous localization and mapping: part I

  • IEEE Robotics & Automation Magazine
  • 2006
VIEW 8 EXCERPTS
HIGHLY INFLUENTIAL

A novel RGB-D SLAM algorithm based on points and plane-patches

  • 2016 IEEE International Conference on Automation Science and Engineering (CASE)
  • 2016
VIEW 1 EXCERPT

Night-time indoor relocalization using depth image with Convolutional Neural Networks

  • 2016 22nd International Conference on Automation and Computing (ICAC)
  • 2016
VIEW 1 EXCERPT