TY - GEN
T1 - GHand: A Graph Convolution Network for 3D Hand Pose Estimation
AU - Wang, Pengsheng
AU - Xue, Guangtao
AU - Li, Ping
AU - Kim, Jinman
AU - Sheng, Bin
AU - Mao, Lijuan
N1 - Funding Information:
Acknowledgments. This work was supported in part by the National Key Research and Development Program of China under Grant 2018YFF0300903, in part by the National Natural Science Foundation of China under Grant 61872241 and Grant 61572316, and in part by the Science and Technology Commission of Shanghai Municipality under Grant 15490503200, Grant 18410750700, Grant 17411952600, and Grant 16DZ0501100.
Publisher Copyright:
© 2020, Springer Nature Switzerland AG.
Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.
PY - 2020/10
Y1 - 2020/10
N2 - Vision-based 3D hand pose estimation plays an important role in the field of human-computer interaction. In recent years, with the development of convolutional neural networks (CNN), the field of 3D hand pose estimation has made a great progress, but there is still a long way to go before the problem is solved. Although recent studies based on CNN networks have greatly improved the recognition accuracy, they usually only pay attention on the regression ability of the network itself, and ignore the structural information of the hands, thus leads to a low accuracy in contrast. In this paper we proposed a new hand pose estimation network, which can fully learn the structural information of hands through an adaptive graph convolutional neural network. The experiment on the public dataset shows the accuracy of our graph convolution network exceeds the SOTA methods in 3D hand pose estimation.
AB - Vision-based 3D hand pose estimation plays an important role in the field of human-computer interaction. In recent years, with the development of convolutional neural networks (CNN), the field of 3D hand pose estimation has made a great progress, but there is still a long way to go before the problem is solved. Although recent studies based on CNN networks have greatly improved the recognition accuracy, they usually only pay attention on the regression ability of the network itself, and ignore the structural information of the hands, thus leads to a low accuracy in contrast. In this paper we proposed a new hand pose estimation network, which can fully learn the structural information of hands through an adaptive graph convolutional neural network. The experiment on the public dataset shows the accuracy of our graph convolution network exceeds the SOTA methods in 3D hand pose estimation.
KW - 3D hand pose estimation
KW - Adaptive graph convolution
KW - Depth image
UR - http://www.scopus.com/inward/record.url?scp=85096517216&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-61864-3_31
DO - 10.1007/978-3-030-61864-3_31
M3 - Conference article published in proceeding or book
AN - SCOPUS:85096517216
SN - 9783030618636
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 374
EP - 381
BT - Advances in Computer Graphics - 37th Computer Graphics International Conference, CGI 2020, Proceedings
A2 - Magnenat-Thalmann, Nadia
A2 - Stephanidis, Constantine
A2 - Papagiannakis, George
A2 - Wu, Enhua
A2 - Thalmann, Daniel
A2 - Sheng, Bin
A2 - Kim, Jinman
A2 - Gavrilova, Marina
PB - Springer Science and Business Media Deutschland GmbH
T2 - 37th Computer Graphics International Conference, CGI 2020
Y2 - 20 October 2020 through 23 October 2020
ER -