Visual Navigation Using Inverse Reinforcement Learning and an Extreme Learning Machine
In this paper, we focus on the challenges of training efficiency, the designation of reward functions, and generalization in reinforcement learning for visual navigation and propose a regularized extreme learning machine-based inverse reinforcement learning approach (RELM-IRL) to improve the navigat...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-08-01
|
Series: | Electronics |
Subjects: | |
Online Access: | https://www.mdpi.com/2079-9292/10/16/1997 |
id |
doaj-35ddbf89de064ad19642a8da0f856bbd |
---|---|
record_format |
Article |
spelling |
doaj-35ddbf89de064ad19642a8da0f856bbd2021-08-26T13:41:47ZengMDPI AGElectronics2079-92922021-08-01101997199710.3390/electronics10161997Visual Navigation Using Inverse Reinforcement Learning and an Extreme Learning MachineQiang Fang0Wenzhuo Zhang1Xitong Wang2College of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, ChinaCollege of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, ChinaCollege of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, ChinaIn this paper, we focus on the challenges of training efficiency, the designation of reward functions, and generalization in reinforcement learning for visual navigation and propose a regularized extreme learning machine-based inverse reinforcement learning approach (RELM-IRL) to improve the navigation performance. Our contributions are mainly three-fold: First, a framework combining extreme learning machine with inverse reinforcement learning is presented. This framework can improve the sample efficiency and obtain the reward function directly from the image information observed by the agent and improve the generation for the new target and the new environment. Second, the extreme learning machine is regularized by multi-response sparse regression and the leave-one-out method, which can further improve the generalization ability. Simulation experiments in the AI-THOR environment showed that the proposed approach outperformed previous end-to-end approaches, thus, demonstrating the effectiveness and efficiency of our approach.https://www.mdpi.com/2079-9292/10/16/1997visual navigationinverse reinforcement learning (IRL)extreme learning machine (ELM)deep learningA3C |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Qiang Fang Wenzhuo Zhang Xitong Wang |
spellingShingle |
Qiang Fang Wenzhuo Zhang Xitong Wang Visual Navigation Using Inverse Reinforcement Learning and an Extreme Learning Machine Electronics visual navigation inverse reinforcement learning (IRL) extreme learning machine (ELM) deep learning A3C |
author_facet |
Qiang Fang Wenzhuo Zhang Xitong Wang |
author_sort |
Qiang Fang |
title |
Visual Navigation Using Inverse Reinforcement Learning and an Extreme Learning Machine |
title_short |
Visual Navigation Using Inverse Reinforcement Learning and an Extreme Learning Machine |
title_full |
Visual Navigation Using Inverse Reinforcement Learning and an Extreme Learning Machine |
title_fullStr |
Visual Navigation Using Inverse Reinforcement Learning and an Extreme Learning Machine |
title_full_unstemmed |
Visual Navigation Using Inverse Reinforcement Learning and an Extreme Learning Machine |
title_sort |
visual navigation using inverse reinforcement learning and an extreme learning machine |
publisher |
MDPI AG |
series |
Electronics |
issn |
2079-9292 |
publishDate |
2021-08-01 |
description |
In this paper, we focus on the challenges of training efficiency, the designation of reward functions, and generalization in reinforcement learning for visual navigation and propose a regularized extreme learning machine-based inverse reinforcement learning approach (RELM-IRL) to improve the navigation performance. Our contributions are mainly three-fold: First, a framework combining extreme learning machine with inverse reinforcement learning is presented. This framework can improve the sample efficiency and obtain the reward function directly from the image information observed by the agent and improve the generation for the new target and the new environment. Second, the extreme learning machine is regularized by multi-response sparse regression and the leave-one-out method, which can further improve the generalization ability. Simulation experiments in the AI-THOR environment showed that the proposed approach outperformed previous end-to-end approaches, thus, demonstrating the effectiveness and efficiency of our approach. |
topic |
visual navigation inverse reinforcement learning (IRL) extreme learning machine (ELM) deep learning A3C |
url |
https://www.mdpi.com/2079-9292/10/16/1997 |
work_keys_str_mv |
AT qiangfang visualnavigationusinginversereinforcementlearningandanextremelearningmachine AT wenzhuozhang visualnavigationusinginversereinforcementlearningandanextremelearningmachine AT xitongwang visualnavigationusinginversereinforcementlearningandanextremelearningmachine |
_version_ |
1721193893550096384 |