Dynamical Deep Generative Latent Modeling of 3D Skeletal Motion

被引:0
|
作者
Amirreza Farnoosh
Sarah Ostadabbas
机构
[1] Northeastern University,Augmented Cognition Lab, Electrical and Computer Engineering Department
来源
关键词
3D skeletal motion; Bayesian inference; Biologically valid interpretation; Generative models; Latent state modeling; Variational inference;
D O I
暂无
中图分类号
学科分类号
摘要
In this paper, we propose a Bayesian switching dynamical model for segmentation of 3D pose data over time that uncovers interpretable patterns in the data and is generative. Our model decomposes highly correlated skeleton data into a set of few spatial basis of switching temporal processes in a low-dimensional latent framework. We parameterize these temporal processes with regard to a switching deep vector autoregressive prior in order to accommodate both multimodal and higher-order nonlinear inter-dependencies. This results in a dynamical deep generative latent model that parses the meaningful intrinsic states in the dynamics of 3D pose data using approximate variational inference, and enables a realistic low-level dynamical generation and segmentation of complex skeleton movements. Our experiments on four biological motion data containing bat flight, salsa dance, walking, and golf datasets substantiate superior performance of our model in comparison with the state-of-the-art methods.
引用
收藏
页码:2695 / 2706
页数:11
相关论文
共 50 条
  • [41] Synthesizing 3D Shapes via Modeling Multi-View Depth Maps and Silhouettes with Deep Generative Networks
    Soltani, Amir Arsalan
    Huang, Haibin
    Wu, Jiajun
    Kulkarni, Tejas D.
    Tenenbaum, Joshua B.
    30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017), 2017, : 2511 - 2519
  • [42] Investigating Pose Representations and Motion Contexts Modeling for 3D Motion Prediction
    Liu, Zhenguang
    Wu, Shuang
    Jin, Shuyuan
    Ji, Shouling
    Liu, Qi
    Lu, Shijian
    Cheng, Li
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2023, 45 (01) : 681 - 697
  • [43] De novo prediction of RNA 3D structures with deep generative models
    Ramakers, Julius
    Blum, Christopher Frederik
    Koenig, Sabrina
    Harmeling, Stefan
    Kollmann, Markus
    PLOS ONE, 2024, 19 (02):
  • [44] On the Value of Using 3D Shape and Electrostatic Similarities in Deep Generative Methods
    Bolcato, Giovanni
    Heid, Esther
    Bostrom, Jonas
    JOURNAL OF CHEMICAL INFORMATION AND MODELING, 2022, 62 (06) : 1388 - 1398
  • [45] Modeling Human Emotions as Reactions to a Dynamical Virtual 3D Face
    Kaminskas, Vytautas
    Vaskevicius, Egidijus
    Vidugiriene, Ausra
    INFORMATICA, 2014, 25 (03) : 425 - 437
  • [46] Motion of an integral curve of a Hamiltonian dynamical system and the evolution equations in 3D
    Bayrakdar, T.
    Ergin, A. A.
    INTERNATIONAL JOURNAL OF GEOMETRIC METHODS IN MODERN PHYSICS, 2017, 14 (12)
  • [47] 3D DYNAMICAL MODEL FOR LIQUID MOTION SIMULATION IN A PARTIALLY FILLED TANK
    Bouazara, M.
    Noui, O.
    Richard, M. J.
    PROCEEDINGS OF THE ASME INTERNATIONAL MECHANICAL ENGINEERING CONGRESS AND EXPOSITION, 2014, VOL 4A, 2015,
  • [48] Learning Latent Representations of 3D Human Pose with Deep Neural Networks
    Isinsu Katircioglu
    Bugra Tekin
    Mathieu Salzmann
    Vincent Lepetit
    Pascal Fua
    International Journal of Computer Vision, 2018, 126 : 1326 - 1341
  • [49] Learning Latent Representations of 3D Human Pose with Deep Neural Networks
    Katircioglu, Isinsu
    Tekin, Bugra
    Salzmann, Mathieu
    Lepetit, Vincent
    Fua, Pascal
    INTERNATIONAL JOURNAL OF COMPUTER VISION, 2018, 126 (12) : 1326 - 1341
  • [50] Human motion segmentation using collaborative representations of 3D skeletal sequences
    Li, Rui
    Liu, Zhenyu
    Tan, Jianrong
    IET COMPUTER VISION, 2018, 12 (04) : 434 - 442