期刊
IEEE SENSORS JOURNAL
卷 21, 期 20, 页码 23174-23184出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JSEN.2021.3107361
关键词
Skeleton; Three-dimensional displays; Kinematics; Sensors; Radar; Pose estimation; Solid modeling; Forward kinematics; 3D human skeleton estimation; radar range-Doppler spectrum; constrained deep learning
资金
- Natural Science Foundation of China [61772574]
- Basic and Applied Basic Research of Guangdong Province, China [2021A1515011758]
This paper explores the incorporation of kinematic constraints into the learning of 3D skeleton reconstruction to eliminate the corruption caused by inaccurate pose estimations. It defines two loss functions to guide the learning process and involves a refined network module to compensate for estimation errors in an iterative manner. Experimental results validate the proposed method.
This paper focuses on the incorporation of kinematic constraints with the learning of 3D skeleton reconstruction without suffering from corrupted pose estimations. Specifically, we devote to exploring a kinematic constrained learning architecture that incorporates the forward kinematics constraint into building a learning model for predicting skeletal key points from observed radar data in the form of range-Doppler spectrum, which contributes to eliminating the corruption of skeleton reconstruction in the scenarios when the radar signals suffer from incomplete sensing process or insufficient granularity due to the bandwidth limitation of radar sensors. In developing our learning paradigm, we define two loss functions, namely the distance loss and the angle loss with respect to the parent-child nodes of skeleton joints, to guide the learning of the deep kinematic network, which is essential to facilitate the skeleton reconstruction without corruption. In addition, the proposed learning architecture involves a refined network module to compensate the estimation offset due to error accumulation of the kinematic model in an iterative fashion. The experimental results are presented to validate the proposed method.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据