Muscle-Aware Retargeting and Null-Space Exploration for Whole-Body Musculoskeletal Control
Burden, Bounty, or Bliss? Muscle-Aware Retargeting and Null-Space Exploration for Whole-Body Musculoskeletal Control
This paper introduces a pipeline that transforms monocular video into a trained whole-body imitation policy controlling 700 muscles. The method addresses kinematic plausibility and high-dimensional redundancy through muscle-aware retargeting with hard constraints and null-space exploration for anatomically structured perturbations. On a 24-clip motion-capture benchmark, the approach solves every reference trajectory, reaches the full horizon with half the samples of a baseline pipeline, and reduces joint-center error by 44% without late collapse.
The authors present a motion-to-policy pipeline combining muscle-aware retargeting and null-space exploration to achieve whole-body musculoskeletal control from human motion data, reporting specific performance metrics on a 24-clip benchmark.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
muscles
700
View original evidence
arXiv:2611.10682v1 Announce Type: new Abstract: Learning whole-body musculoskeletal control from human motion is important for developing human-like robots, assistive devices, and rehabilitation systems. This task presents two main challenges, namely reference trajectories that are kinematically plausible but difficult to reproduce through muscle-driven dynamics, and a highly redundant 700-dimensional action space in
Open source S2clips
24
Reported trials
View original evidence
s affect muscle-driven tracking. Second, null-space exploration (NSE) estimates a one-control-frame mapping from muscle excitations to generalized forces and injects bounded, anatomically structured perturbations along its null space into the actions of a value-guided flow policy. On a 24-clip motion-capture benchmark, MAR and NSE together solve every reference, reach the full horizon with half the samples of the bas
Open source S4
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
arXiv:2611.10682v1 Announce Type: new Abstract: Learning whole-body musculoskeletal control from human motion is important for developing human-like robots, assistive devices, and rehabilitation systems. This task presents two main challenges, namely reference trajectories that are kinematically plausible but difficult to reproduce through muscle-driven dynamics, and a highly redundant 700-dimensional action space in
Open source S2
s affect muscle-driven tracking. Second, null-space exploration (NSE) estimates a one-control-frame mapping from muscle excitations to generalized forces and injects bounded, anatomically structured perturbations along its null space into the actions of a value-guided flow policy. On a 24-clip motion-capture benchmark, MAR and NSE together solve every reference, reach the full horizon with half the samples of the bas
Open source S4
eline pipeline, and lower joint-center error by 44% without late collapse. To the best of our knowledge, this is the first end-to-end pipeline that transforms monocular video into a trained whole-body imitation policy controlling 700 muscles.
Open source S5
Burden, Bounty, or Bliss? Muscle-Aware Retargeting and Null-Space Exploration for Whole-Body Musculoskeletal Control
Open source S1
which unstructured exploration can easily disrupt the target motion. In this paper, we propose a motion-to-policy pipeline that exploits muscle redundancy in two complementary stages. First, muscle-aware retargeting (MAR) performs bounded adaptation of kinematic trajectories subject to hard semantic, anatomical, and contact constraints. We motivate the constraint bounds by analyzing how reference-trajectory artifact
Open source S3
Source：arXiv Robotics — research abstracts · arxiv.org