Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Ma, Shuhao, Huang, Zeyi, Cao, Yu, Doorsamy, Wesley, Shi, Chaoyang, Li, Jun, Zhang, Zhi-Qiang
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2511.10878
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918201631178752
author Ma, Shuhao
Huang, Zeyi
Cao, Yu
Doorsamy, Wesley
Shi, Chaoyang
Li, Jun
Zhang, Zhi-Qiang
author_facet Ma, Shuhao
Huang, Zeyi
Cao, Yu
Doorsamy, Wesley
Shi, Chaoyang
Li, Jun
Zhang, Zhi-Qiang
contents Time-efficient estimation of muscle activations and forces across multi-joint systems is critical for clinical assessment and assistive device control. However, conventional approaches are computationally expensive and lack a high-quality labeled dataset for multi-joint applications. To address these challenges, we propose a physics-informed deep learning framework that estimates muscle activations and forces directly from kinematics. The framework employs a novel Multi-Joint Cross-Attention (MJCA) module with Bidirectional Gated Recurrent Unit (BiGRU) layers to capture inter-joint coordination, enabling each joint to adaptively integrate motion information from others. By embedding multi-joint dynamics, inter-joint coupling, and external force interactions into the loss function, our Physics-Informed MJCA-BiGRU (PI-MJCA-BiGRU) delivers physiologically consistent predictions without labeled data while enabling time-efficient inference. Experimental validation on two datasets demonstrates that PI-MJCA-BiGRU achieves performance comparable to conventional supervised methods without requiring ground-truth labels, while the MJCA module significantly enhances inter-joint coordination modeling compared to other baseline architectures.
format Preprint
id arxiv_https___arxiv_org_abs_2511_10878
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Multi-Joint Physics-Informed Deep Learning Framework for Time-Efficient Inverse Dynamics
Ma, Shuhao
Huang, Zeyi
Cao, Yu
Doorsamy, Wesley
Shi, Chaoyang
Li, Jun
Zhang, Zhi-Qiang
Machine Learning
Human-Computer Interaction
Signal Processing
Time-efficient estimation of muscle activations and forces across multi-joint systems is critical for clinical assessment and assistive device control. However, conventional approaches are computationally expensive and lack a high-quality labeled dataset for multi-joint applications. To address these challenges, we propose a physics-informed deep learning framework that estimates muscle activations and forces directly from kinematics. The framework employs a novel Multi-Joint Cross-Attention (MJCA) module with Bidirectional Gated Recurrent Unit (BiGRU) layers to capture inter-joint coordination, enabling each joint to adaptively integrate motion information from others. By embedding multi-joint dynamics, inter-joint coupling, and external force interactions into the loss function, our Physics-Informed MJCA-BiGRU (PI-MJCA-BiGRU) delivers physiologically consistent predictions without labeled data while enabling time-efficient inference. Experimental validation on two datasets demonstrates that PI-MJCA-BiGRU achieves performance comparable to conventional supervised methods without requiring ground-truth labels, while the MJCA module significantly enhances inter-joint coordination modeling compared to other baseline architectures.
title Multi-Joint Physics-Informed Deep Learning Framework for Time-Efficient Inverse Dynamics
topic Machine Learning
Human-Computer Interaction
Signal Processing
url https://arxiv.org/abs/2511.10878