Embodiment-aware control by inference over the operator: a simulation study
Embodiment-aware control by inference over the operator: a simulation study
This research introduces an embodiment-aware controller that infers operator embodiment and visuo-proprioceptive cue weighting from implicit gaze and pupil signals using a discrete Active Inference agent. In simulations with heterogeneous synthetic operators, the method increased embodiment scores from 0.76 to 1.66 on a 0-2 scale, reaching within 0.02 of an oracle across ten random seeds. Unlike prediction-error minimizers that withhold feedback or fixed-setting controllers, this approach adapts to operator needs while respecting explicit preferences. The model also reproduced known embodiment-performance relations in human data and developed embodiment on plausible time scales. Results indicate that the benefit stems from inference over the operator with explicit preferences rather than information seeking alone.
The study demonstrates that inferring operator embodiment and cue weighting from implicit gaze and pupil signals improves teleoperation performance in simulation, achieving near-oracle results across random seeds while highlighting the need for calibrated human data before deployment.
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.
- Environment
- SimulationOpen source S1
Reported numbers
embodiment score
0.76–1.66
View original evidence
tive. We formulate an embodiment-aware controller that instead infers the operator's embodiment and visuo-proprioceptive cue weighting from implicit gaze and pupil signals and task outcome, and chooses bounded device settings under explicit preferences, cast as a discrete Active Inference agent. In simulations with heterogeneous synthetic operators, it raised embodiment from 0.76 to 1.66 on a 0-2 scale over the best
Open source S3
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
Embodiment-aware control by inference over the operator: a simulation study
Open source S1
tive. We formulate an embodiment-aware controller that instead infers the operator's embodiment and visuo-proprioceptive cue weighting from implicit gaze and pupil signals and task outcome, and chooses bounded device settings under explicit preferences, cast as a discrete Active Inference agent. In simulations with heterogeneous synthetic operators, it raised embodiment from 0.76 to 1.66 on a 0-2 scale over the best
Open source S3
Source：arXiv Robotics — research abstracts · arxiv.org