Full Text:   <36>

CLC number: 

On-line Access: 2022-08-18

Received: 2022-03-20

Revision Accepted: 2022-07-29

Crosschecked: 0000-00-00

Cited: 0

Clicked: 95

Citations:  Bibtex RefMan EndNote GB/T7714

-   Go to

Article info.
Open peer comments

Journal of Zhejiang University SCIENCE C 1998 Vol.-1 No.-1 P.

http://doi.org/10.1631/FITEE.2200109


A deep Q-learning network-based active object detection model with a novel training algorithm for service robots


Author(s):  Shaopeng LIU, Guohui TIAN, Yongcheng CUI, Xuyang SHAO

Affiliation(s):  School of Control Science and Engineering, Shandong University, Jinan 250061, China

Corresponding email(s):   shaopeng.liu66@mail.sdu.edu.cn, g.h.tian@sdu.edu.cn, cuiyc@mail.sdu.edu.cn, 201834562@mail.sdu.edu.cn

Key Words:  Active object detection, Deep Q-learning Network, Training method, Service robots


Shaopeng LIU, Guohui TIAN, Yongcheng CUI, Xuyang SHAO. A deep Q-learning network-based active object detection model with a novel training algorithm for service robots[J]. Frontiers of Information Technology & Electronic Engineering, 1998, -1(-1): .

@article{title="A deep Q-learning network-based active object detection model with a novel training algorithm for service robots",
author="Shaopeng LIU, Guohui TIAN, Yongcheng CUI, Xuyang SHAO",
journal="Frontiers of Information Technology & Electronic Engineering",
volume="-1",
number="-1",
pages="",
year="1998",
publisher="Zhejiang University Press & Springer",
doi="10.1631/FITEE.2200109"
}

%0 Journal Article
%T A deep Q-learning network-based active object detection model with a novel training algorithm for service robots
%A Shaopeng LIU
%A Guohui TIAN
%A Yongcheng CUI
%A Xuyang SHAO
%J Journal of Zhejiang University SCIENCE C
%V -1
%N -1
%P
%@ 2095-9184
%D 1998
%I Zhejiang University Press & Springer
%DOI 10.1631/FITEE.2200109

TY - JOUR
T1 - A deep Q-learning network-based active object detection model with a novel training algorithm for service robots
A1 - Shaopeng LIU
A1 - Guohui TIAN
A1 - Yongcheng CUI
A1 - Xuyang SHAO
J0 - Journal of Zhejiang University Science C
VL - -1
IS - -1
SP -
EP -
%@ 2095-9184
Y1 - 1998
PB - Zhejiang University Press & Springer
ER -
DOI - 10.1631/FITEE.2200109


Abstract: 
This paper focuses on the problem of active object detection (AOD). AOD is important for the service robot to complete tasks in the family environment, and leads the robot in approaching the target object by taking appropriate moving actions. Most of the current AOD methods are based on reinforcement learning with low training efficiency and testing accuracy. Therefore, an AOD model based on a deep Q-learning Network (DQN) with a novel training algorithm is proposed in this paper. The DQN model is designed to fit the Q-value of various actions, and includes state space, feature extraction, and a multilayer perceptron. In contrast to existing research, a novel training algorithm based on memory is designed for the proposed DQN model to improve training efficiency and test accuracy. In addition, a method of generating the end state is presented to judge when to stop the AOD task during the training process. Sufficient comparative experiments and ablation studies were performed in an AOD dataset, which proves that the presented method has better performance than the comparable methods and the proposed training algorithm is more effective than the raw training algorithm.

Darkslateblue:Affiliate; Royal Blue:Author; Turquoise:Article

Open peer comments: Debate/Discuss/Question/Opinion

<1>

Please provide your name, email address and a comment





Journal of Zhejiang University-SCIENCE, 38 Zheda Road, Hangzhou 310027, China
Tel: +86-571-87952783; E-mail: cjzhang@zju.edu.cn
Copyright © 2000 - 2022 Journal of Zhejiang University-SCIENCE