Home > Research > Publications & Outputs > Robust Reinforcement Learning Based Visual Serv...

Research

Associated organisational units

Links

Text available via DOI:

Keywords

View graph of relations

Robust Reinforcement Learning Based Visual Servoing with Convolutional Features

Research output: Contribution to Journal/Magazine › Conference article › peer-review

Published

Standard

Robust Reinforcement Learning Based Visual Servoing with Convolutional Features. / Fei, Haolin; Wang, Ziwei; Kennedy, Andrew.
In: IFAC-PapersOnLine, Vol. 56, No. 2, 31.12.2023, p. 9781-9786.

Research output: Contribution to Journal/Magazine › Conference article › peer-review

Harvard

Fei, H, Wang, Z & Kennedy, A 2023, 'Robust Reinforcement Learning Based Visual Servoing with Convolutional Features', IFAC-PapersOnLine, vol. 56, no. 2, pp. 9781-9786. https://doi.org/10.1016/j.ifacol.2023.10.295

APA

Fei, H., Wang, Z., & Kennedy, A. (2023). Robust Reinforcement Learning Based Visual Servoing with Convolutional Features. IFAC-PapersOnLine, 56(2), 9781-9786. https://doi.org/10.1016/j.ifacol.2023.10.295

Vancouver

Fei H, Wang Z, Kennedy A. Robust Reinforcement Learning Based Visual Servoing with Convolutional Features. IFAC-PapersOnLine. 2023 Dec 31;56(2):9781-9786. doi: 10.1016/j.ifacol.2023.10.295

Author

Fei, Haolin ; Wang, Ziwei ; Kennedy, Andrew. / Robust Reinforcement Learning Based Visual Servoing with Convolutional Features. In: IFAC-PapersOnLine. 2023 ; Vol. 56, No. 2. pp. 9781-9786.

Bibtex

@article{7424f7b50fe846edaeccdefee070eff1,
title = "Robust Reinforcement Learning Based Visual Servoing with Convolutional Features",
abstract = "Image-based visual servoing is challenging as the robot needs to locate the object and learn to control the arm in the image plane, which often undergo significant interference such as ambient light, distractions, and background clutter. Recent studies shows that the control policy can be efficiently learned by the reinforcement learning. In this paper, we present a data-driven image-based closed-loop visual servoing method via reinforcement learning algorithm without any prior knowledge of the task object or the intrinsic camera parameters. We first locate the object with a convolutional neural network backbone feature extraction network. Moreover the robot can determine the relative motion of the camera and servo the camera to the desired pose. We demonstrate that the reinforcement learning based approach is capable of steering the camera with only a single template image of the task object.",
keywords = "Reinforcement Leaning, Visual Servoing, Imitation Learning",
author = "Haolin Fei and Ziwei Wang and Andrew Kennedy",
year = "2023",
month = dec,
day = "31",
doi = "10.1016/j.ifacol.2023.10.295",
language = "English",
volume = "56",
pages = "9781--9786",
journal = "IFAC-PapersOnLine",
issn = "2405-8963",
publisher = "IFAC Secretariat",
number = "2",

}

RIS

TY - JOUR

T1 - Robust Reinforcement Learning Based Visual Servoing with Convolutional Features

AU - Fei, Haolin

AU - Wang, Ziwei

AU - Kennedy, Andrew

PY - 2023/12/31

Y1 - 2023/12/31

N2 - Image-based visual servoing is challenging as the robot needs to locate the object and learn to control the arm in the image plane, which often undergo significant interference such as ambient light, distractions, and background clutter. Recent studies shows that the control policy can be efficiently learned by the reinforcement learning. In this paper, we present a data-driven image-based closed-loop visual servoing method via reinforcement learning algorithm without any prior knowledge of the task object or the intrinsic camera parameters. We first locate the object with a convolutional neural network backbone feature extraction network. Moreover the robot can determine the relative motion of the camera and servo the camera to the desired pose. We demonstrate that the reinforcement learning based approach is capable of steering the camera with only a single template image of the task object.

AB - Image-based visual servoing is challenging as the robot needs to locate the object and learn to control the arm in the image plane, which often undergo significant interference such as ambient light, distractions, and background clutter. Recent studies shows that the control policy can be efficiently learned by the reinforcement learning. In this paper, we present a data-driven image-based closed-loop visual servoing method via reinforcement learning algorithm without any prior knowledge of the task object or the intrinsic camera parameters. We first locate the object with a convolutional neural network backbone feature extraction network. Moreover the robot can determine the relative motion of the camera and servo the camera to the desired pose. We demonstrate that the reinforcement learning based approach is capable of steering the camera with only a single template image of the task object.

KW - Reinforcement Leaning

KW - Visual Servoing

KW - Imitation Learning

U2 - 10.1016/j.ifacol.2023.10.295

DO - 10.1016/j.ifacol.2023.10.295

M3 - Conference article

VL - 56

SP - 9781

EP - 9786

JO - IFAC-PapersOnLine

JF - IFAC-PapersOnLine

SN - 2405-8963

IS - 2

ER -