Domain Adaptation for Imitation Learning Using Generative Adversarial Network
Imitation learning is an effective approach for an autonomous agent to learn control policies when an explicit reward function is unavailable, using demonstrations provided from an expert. However, standard imitation learning methods assume that the agents and the demonstrations provided by the expe...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-07-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/21/14/4718 |
id |
doaj-948bdeb9c5c5496ba84e76dd68c59873 |
---|---|
record_format |
Article |
spelling |
doaj-948bdeb9c5c5496ba84e76dd68c598732021-07-23T14:05:27ZengMDPI AGSensors1424-82202021-07-01214718471810.3390/s21144718Domain Adaptation for Imitation Learning Using Generative Adversarial NetworkTho Nguyen Duc0Chanh Minh Tran1Phan Xuan Tan2Eiji Kamioka3Graduate School of Engineering and Science, Shibaura Institute of Technology, Tokyo 135-8548, JapanGraduate School of Engineering and Science, Shibaura Institute of Technology, Tokyo 135-8548, JapanDepartment of Information and Communications Engineering, Shibaura Institute of Technology, Tokyo 135-8548, JapanGraduate School of Engineering and Science, Shibaura Institute of Technology, Tokyo 135-8548, JapanImitation learning is an effective approach for an autonomous agent to learn control policies when an explicit reward function is unavailable, using demonstrations provided from an expert. However, standard imitation learning methods assume that the agents and the demonstrations provided by the expert are in the same domain configuration. Such an assumption has made the learned policies difficult to apply in another distinct domain. The problem is formalized as domain adaptive imitation learning, which is the process of learning how to perform a task optimally in a learner domain, given demonstrations of the task in a distinct expert domain. We address the problem by proposing a model based on Generative Adversarial Network. The model aims to learn both domain-shared and domain-specific features and utilizes it to find an optimal policy across domains. The experimental results show the effectiveness of our model in a number of tasks ranging from low to complex high-dimensional.https://www.mdpi.com/1424-8220/21/14/4718imitation learningdomain adaptive imitation learninggenerative adversarial network |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Tho Nguyen Duc Chanh Minh Tran Phan Xuan Tan Eiji Kamioka |
spellingShingle |
Tho Nguyen Duc Chanh Minh Tran Phan Xuan Tan Eiji Kamioka Domain Adaptation for Imitation Learning Using Generative Adversarial Network Sensors imitation learning domain adaptive imitation learning generative adversarial network |
author_facet |
Tho Nguyen Duc Chanh Minh Tran Phan Xuan Tan Eiji Kamioka |
author_sort |
Tho Nguyen Duc |
title |
Domain Adaptation for Imitation Learning Using Generative Adversarial Network |
title_short |
Domain Adaptation for Imitation Learning Using Generative Adversarial Network |
title_full |
Domain Adaptation for Imitation Learning Using Generative Adversarial Network |
title_fullStr |
Domain Adaptation for Imitation Learning Using Generative Adversarial Network |
title_full_unstemmed |
Domain Adaptation for Imitation Learning Using Generative Adversarial Network |
title_sort |
domain adaptation for imitation learning using generative adversarial network |
publisher |
MDPI AG |
series |
Sensors |
issn |
1424-8220 |
publishDate |
2021-07-01 |
description |
Imitation learning is an effective approach for an autonomous agent to learn control policies when an explicit reward function is unavailable, using demonstrations provided from an expert. However, standard imitation learning methods assume that the agents and the demonstrations provided by the expert are in the same domain configuration. Such an assumption has made the learned policies difficult to apply in another distinct domain. The problem is formalized as domain adaptive imitation learning, which is the process of learning how to perform a task optimally in a learner domain, given demonstrations of the task in a distinct expert domain. We address the problem by proposing a model based on Generative Adversarial Network. The model aims to learn both domain-shared and domain-specific features and utilizes it to find an optimal policy across domains. The experimental results show the effectiveness of our model in a number of tasks ranging from low to complex high-dimensional. |
topic |
imitation learning domain adaptive imitation learning generative adversarial network |
url |
https://www.mdpi.com/1424-8220/21/14/4718 |
work_keys_str_mv |
AT thonguyenduc domainadaptationforimitationlearningusinggenerativeadversarialnetwork AT chanhminhtran domainadaptationforimitationlearningusinggenerativeadversarialnetwork AT phanxuantan domainadaptationforimitationlearningusinggenerativeadversarialnetwork AT eijikamioka domainadaptationforimitationlearningusinggenerativeadversarialnetwork |
_version_ |
1721286003530924032 |