POMDP-based action planning for the recognition of occluded objects with Humanoid robots

被引：0
|
作者：
Tsuru, Masato [1 ]
Gergondet, Pierre [2 ]
Motoda, Tomohiro [1 ]
Escande, Adrien [2 ]
Yoshida, Eiichi [3 ]
Ramirez-Alpizar, Ixchel G. [4 ]
Wan, Weiwei [1 ]
Harada, Kensuke [5 ]
机构：
[1] Osaka Univ, Grad Sch Engn Sci, 1-3 Machikaneyama, Osaka, Japan
[2] AIST, CNRS JRL, 1-1-1 Umezono, Tsukuba, Ibaraki, Japan
[3] AIST, Intelligent Syst Res Inst, 1-1-1 Umezono, Tsukuba, Ibaraki, Japan
[4] AIST, AIRC, Koto Ku, 2-3-26 Aomi, Tokyo, Japan
[5] Osaka Univ, Grad Sch Engn Sci, AIRC, AIST, Osaka, Japan
来源：
PROCEEDINGS OF THE 2020 INTERNATIONAL CONFERENCE ON ARTIFICIAL LIFE AND ROBOTICS (ICAROB2020) | 2020年
关键词：
humanoid robot; POMDP; planning; point cloud; occlusion; 6DoF registration;
D O I：
暂无
中图分类号：
TP18 [人工智能理论];
学科分类号：
081104 ; 0812 ; 0835 ; 1405 ;
摘要：
In this paper, we present a high-layer motion planner which plans humanoid robot actions to search for object models in the robot workspace. To overcome the occlusion problem, our proposed method plans to get different perspectives of the object. POMDP (Partially Observable Markov Decision Process) is used to determine the observation pose of a robot. The planner then builds a comprehensive point cloud of by merging the point clouds gathered from different positions. The point cloud is compared to a 3D model of the object to estimate the pose of the real object.
引用
收藏
页码：281 / 284
页数：4
相关论文
共 50 条
  • [1] αPOMDP: POMDP-based user-adaptive decision-making for social robots
    Martins, Goncalo S.
    Al Tair, Hend
    Santos, Luis
    Dias, Jorge
    PATTERN RECOGNITION LETTERS, 2019, 118 : 94 - 103
  • [2] Non-linearity Measure for POMDP-based Motion Planning
    Hoerger, Marcus
    Kurniawati, Hanna
    Elfes, Alberto
    INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH, 2024, 43 (10): : 1629 - 1646
  • [3] POMDP-based online target detection and recognition for autonomous UAVs
    Chanel, Caroline P. Carvalho
    Teichteil-Koenigsbuch, Florent
    Lesire, Charles
    20TH EUROPEAN CONFERENCE ON ARTIFICIAL INTELLIGENCE (ECAI 2012), 2012, 242 : 955 - +
  • [4] Virtual tactile POMDP-based path planning for object localization and grasping
    Wang, Chao
    Zang, Xizhe
    Song, Chao
    Liu, Zhiyang
    Zhao, Jie
    Ang Jr, Marcelo H.
    MEASUREMENT, 2024, 230
  • [5] POMDP-based probabilistic decision making for path planning in wheeled mobile robot
    Deshpande S.V.
    R H.
    Walambe R.
    Cognitive Robotics, 2024, 4 : 104 - 115
  • [6] Intelligent Facial Action and Emotion Recognition for Humanoid Robots
    Zhang, Li
    Hossain, Alamgir
    Jiang, Ming
    PROCEEDINGS OF THE 2014 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2014, : 739 - 746
  • [7] Vision based motion planning of humanoid robots
    Ercan, MF
    Zhou, CJ
    ROBOTIC WELDING, INTELLIGENCE AND AUTOMATION, 2004, 299 : 212 - 227
  • [8] POMDP Based Action Planning and Human Error Detection
    Jean-Baptiste, Emilie M. D.
    Rotshtein, Pia
    Russell, Martin
    ARTIFICIAL INTELLIGENCE APPLICATIONS AND INNOVATIONS, 2015, 458 : 250 - 265
  • [9] Walking gaits planning for humanoid robots based on HMCD
    Peng, Xiu-Yan
    Ding, Hu
    Qu, Yang
    Xitong Fangzhen Xuebao / Journal of System Simulation, 2008, 20 (24): : 6834 - 6838
  • [10] POMDP-Based Real-Time Path Planning for Manipulation of Multiple Microparticles via Optoelectronic Tweezers
    Liu, Jiaxin
    Wang, Huaping
    Liu, Menghua
    Zhao, Ran
    Zhao, Yanfeng
    Sun, Tao
    Shi, Qing
    CYBORG AND BIONIC SYSTEMS, 2022, 2022
12345