SIA OpenIR  > 空间自动化技术研究室
Active Object Detection Using Double DQN and Prioritized Experience Replay
Han XN(韩小宁)1; Liu H(刘华平)2; Sun FC(孙富春)2; Yang, Dongfang3
作者部门空间自动化技术研究室
会议名称2018 International Joint Conference on Neural Networks, IJCNN 2018
会议日期July 8-13, 2018
会议地点Rio de Janeiro, Brazil
会议录名称Proceedings of the International Joint Conference on Neural Networks
出版者IEEE
出版地New York
2018
页码1-7
收录类别EI
EI收录号20184706086233
产权排序1
ISBN号978-1-5090-6014-6
摘要Visual object detection is one of the fundamental tasks in computer vision and robotics. Small scale, partial capture and occlusion often occur in robotic applications, most existing object detection algorithms perform poorly in such situations. While a robot can look at one object from different views and plan its trajectory in the next few steps, which can lead to better observations. We formulate it as a sequential action-decision process, and develop a deep reinforcement learning architecture to solve the active object detection problem. A double deep Q-learning network (DQN) is applied to predict an action at each step. Experimental validation on the Active Vision Dataset shows the efficiency of the proposed method.
语种英语
文献类型会议论文
条目标识符http://ir.sia.cn/handle/173321/23592
专题空间自动化技术研究室
通讯作者Han XN(韩小宁)
作者单位1.Chinese Academy of Sciences, Shenyang Institute of Automation, Shenyang, China
2.Department of Computer Science and Technology, Tsinghua University, Beijing, China
3.Xi'An High Tech Research Institution, China
推荐引用方式
GB/T 7714
Han XN,Liu H,Sun FC,et al. Active Object Detection Using Double DQN and Prioritized Experience Replay[C]. New York:IEEE,2018:1-7.
条目包含的文件 下载所有文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
Active Object Detect(9517KB)会议论文 开放获取CC BY-NC-SA浏览 下载
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Han XN(韩小宁)]的文章
[Liu H(刘华平)]的文章
[Sun FC(孙富春)]的文章
百度学术
百度学术中相似的文章
[Han XN(韩小宁)]的文章
[Liu H(刘华平)]的文章
[Sun FC(孙富春)]的文章
必应学术
必应学术中相似的文章
[Han XN(韩小宁)]的文章
[Liu H(刘华平)]的文章
[Sun FC(孙富春)]的文章
相关权益政策
暂无数据
收藏/分享
文件名: Active Object Detection Using Double DQN and Prioritized Experience Replay.pdf
格式: Adobe PDF
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。