Open Access
Open access
IEEE Access, volume 8, pages 195608-195621

Real-Time Object Navigation With Deep Neural Networks and Hierarchical Reinforcement Learning

Publication typeJournal Article
Publication date2020-10-28
Journal: IEEE Access
Quartile SCImago
Q1
Quartile WOS
Q2
Impact factor3.9
ISSN21693536
General Materials Science
General Engineering
General Computer Science
Abstract
In the last years, deep learning and reinforcement learning methods have significantly improved mobile robots in such fields as perception, navigation, and planning. But there are still gaps in applying these methods to real robots due to the low computational efficiency of recent neural network architectures and their poor adaptability to robotic experiments’ realities. In this article, we consider an important task in mobile robotics - navigation to an object using an RGB-D camera. We develop a new neural network framework for robot control that is fast and resistant to possible noise in sensors and actuators. We propose an original integration of semantic segmentation, mapping, localization, and reinforcement learning methods to improve the effectiveness of exploring the environment, finding the desired object, and quickly navigating to it. We created a new HISNav dataset based on the Habitat virtual environment, which allowed us to use simulation experiments to pre-train the model and then upload it to a real robot. Our architecture is adapted to work in a real-time environment and fully implements modern trends in this area.

Citations by journals

1
2
3
Lecture Notes in Networks and Systems
Lecture Notes in Networks and Systems, 3, 11.11%
Lecture Notes in Networks and Systems
3 publications, 11.11%
Lecture Notes in Computer Science
Lecture Notes in Computer Science, 2, 7.41%
Lecture Notes in Computer Science
2 publications, 7.41%
Procedia Computer Science
Procedia Computer Science, 1, 3.7%
Procedia Computer Science
1 publication, 3.7%
Journal of Physics: Conference Series
Journal of Physics: Conference Series, 1, 3.7%
Journal of Physics: Conference Series
1 publication, 3.7%
Studies in Computational Intelligence
Studies in Computational Intelligence, 1, 3.7%
Studies in Computational Intelligence
1 publication, 3.7%
SoftwareX
SoftwareX, 1, 3.7%
SoftwareX
1 publication, 3.7%
Automation and Remote Control
Automation and Remote Control, 1, 3.7%
Automation and Remote Control
1 publication, 3.7%
IEEE Robotics and Automation Letters
IEEE Robotics and Automation Letters, 1, 3.7%
IEEE Robotics and Automation Letters
1 publication, 3.7%
Journal of Real-Time Image Processing
Journal of Real-Time Image Processing, 1, 3.7%
Journal of Real-Time Image Processing
1 publication, 3.7%
Wireless Personal Communications
Wireless Personal Communications, 1, 3.7%
Wireless Personal Communications
1 publication, 3.7%
Robotics
Robotics, 1, 3.7%
Robotics
1 publication, 3.7%
Advances in Computational Intelligence and Robotics
Advances in Computational Intelligence and Robotics, 1, 3.7%
Advances in Computational Intelligence and Robotics
1 publication, 3.7%
International Journal of Geographical Information Science
International Journal of Geographical Information Science, 1, 3.7%
International Journal of Geographical Information Science
1 publication, 3.7%
Multimedia Tools and Applications
Multimedia Tools and Applications, 1, 3.7%
Multimedia Tools and Applications
1 publication, 3.7%
1
2
3

Citations by publishers

1
2
3
4
5
6
7
8
9
Springer Nature
Springer Nature, 9, 33.33%
Springer Nature
9 publications, 33.33%
Elsevier
Elsevier, 2, 7.41%
Elsevier
2 publications, 7.41%
IOP Publishing
IOP Publishing, 1, 3.7%
IOP Publishing
1 publication, 3.7%
Pleiades Publishing
Pleiades Publishing, 1, 3.7%
Pleiades Publishing
1 publication, 3.7%
IEEE
IEEE, 1, 3.7%
IEEE
1 publication, 3.7%
Multidisciplinary Digital Publishing Institute (MDPI)
Multidisciplinary Digital Publishing Institute (MDPI), 1, 3.7%
Multidisciplinary Digital Publishing Institute (MDPI)
1 publication, 3.7%
IGI Global
IGI Global, 1, 3.7%
IGI Global
1 publication, 3.7%
Taylor & Francis
Taylor & Francis, 1, 3.7%
Taylor & Francis
1 publication, 3.7%
1
2
3
4
5
6
7
8
9
  • We do not take into account publications that without a DOI.
  • Statistics recalculated only for publications connected to researchers, organizations and labs registered on the platform.
  • Statistics recalculated weekly.
Metrics
Share
Cite this
GOST |
Cite this
GOST Copy
Staroverov A. et al. Real-Time Object Navigation With Deep Neural Networks and Hierarchical Reinforcement Learning // IEEE Access. 2020. Vol. 8. pp. 195608-195621.
GOST all authors (up to 50) Copy
Staroverov A., Yudin D., Belkin I., Adeshkin V., Solomentsev Y. K., Panov A. Real-Time Object Navigation With Deep Neural Networks and Hierarchical Reinforcement Learning // IEEE Access. 2020. Vol. 8. pp. 195608-195621.
RIS |
Cite this
RIS Copy
TY - JOUR
DO - 10.1109/ACCESS.2020.3034524
UR - https://doi.org/10.1109%2FACCESS.2020.3034524
TI - Real-Time Object Navigation With Deep Neural Networks and Hierarchical Reinforcement Learning
T2 - IEEE Access
AU - Staroverov, Aleksei
AU - Yudin, D.
AU - Belkin, Ilya
AU - Adeshkin, Vasily
AU - Solomentsev, Yaroslav K
AU - Panov, Aleksandr
PY - 2020
DA - 2020/10/28 00:00:00
PB - IEEE
SP - 195608-195621
VL - 8
SN - 2169-3536
ER -
BibTex
Cite this
BibTex Copy
@article{2020_Staroverov
author = {Aleksei Staroverov and D. Yudin and Ilya Belkin and Vasily Adeshkin and Yaroslav K Solomentsev and Aleksandr Panov},
title = {Real-Time Object Navigation With Deep Neural Networks and Hierarchical Reinforcement Learning},
journal = {IEEE Access},
year = {2020},
volume = {8},
publisher = {IEEE},
month = {oct},
url = {https://doi.org/10.1109%2FACCESS.2020.3034524},
pages = {195608--195621},
doi = {10.1109/ACCESS.2020.3034524}
}
Found error?