Zusammenfassung
This dissertation deals with machine learning techniques for inverse dynamics of human motion. Inverse
dynamics refers to the derivation of acting forces and moments from the motion of a kinematic model. More precisely, the objective is to estimate joint torques, ground reaction forces and ground reaction moments at both feet based on the three-dimensional input motion of a skeletal model. The problem is solved using a data-driven machine learning approach, proposing several regression models that are particularly suitable with respect to limited data availability. The goal is to exploit the inherent strengths of machine learning, such as fast and noiseresistant data analysis. The described methods are able to predict underlying joint torques and exterior forces with high precision (on gait sequences: relative root mean squared errors of 7.0 %, 16.1 % and 11.9 % for reaction forces, reaction moments and joint moments which correspond to Pearson‘s correlation coefficients of 0.91, 0.83 and 0.82), while reducing computation times by two orders of magnitude compared to traditional optimization.
Contents
1 Introduction 1
1.1 Applications and Challenges of Inverse Dynamics . . ....
Schlagworte
Machine Learning Maschinelles Lernen Künstliche Neuronale Netze inverse Dynamik menschliche Bewegung Gelenkmomente Ganganalyse selbstüberwachtes Lernen inverse dynamics human motion joint moments gait analysis artificial neural networks self-supervised learning- 1–14 1 Introduction 1–14
- 15–22 2 Related work 15–22
- 23–61 3 Fundamentals 23–61
- 123–126 7 Conclusions 123–126
- 127–129 a Appendix 127–129
- 130–152 Bibliography 130–152
4 Treffer gefunden
- „... . “Weakly-Supervised Discovery of Geometry-Aware Representation for 3D Human PoseEstimation.” In: Proceedings of the ...” „... Learning for Bottom-Up HumanPose Estimation.” In: Proceedings of the IEEE/CVF Conference on Computer ...” „... , 2008.[50] Rıza Alp Güler, Natalia Neverova, and Iasonas Kokkinos. “DensePose: DenseHuman Pose Estimation in ...”
- „... correspondences on the human model (in canonical shapeand pose) and a forward map parameterized by the human ...” „... reduced supervision in the broader field of human motion analysis(2.3).2.1 inverse dynamics by physical ...” „... simulationThe study of human movement has a very long tradition in human history. As a formof art it already ...”
- „... pose estimation [19, 50,90, 137, 162]. The regression of acting forces and moments from an observed ...” „... learning of human dynamics is presented. The application ofmachine learning to inverse dynamics is ...” „... motivated by the tremendous success of artificialneural networks in related problems such as 2D and 3D human ...”
- „... 3F U N DA M E N TA L S3.1 rigid body motionDue to the stiffness of skeletal bones, the human body ...” „... (3). In order to mathematically describe the human body as a system of rigidbodies, the following ...” „... representation of rigid motion for individual bodies. Inorder to model the motion of an entire human skeleton, we ...”