A self-supervised PINN estimates human joint kinematics and kinetics from sparse IMU data by enforcing physical plausibility via a body model, reporting RMSD of 8.7 deg and 4.9 BWBH% on walking and running.
Kingma and Jimmy Ba
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SSPINNpose: A Self-Supervised PINN for Inertial Pose and Dynamics Estimation
A self-supervised PINN estimates human joint kinematics and kinetics from sparse IMU data by enforcing physical plausibility via a body model, reporting RMSD of 8.7 deg and 4.9 BWBH% on walking and running.