An RNN arbitration model is trained on user data collected under shared control to blend human and robot commands in teleoperated pick-and-place tasks, with preliminary virtual-environment results compared to handcrafted baselines.
A differentiable policy for shared autonomy
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning Arbitration for Shared Autonomy by Hindsight Data Aggregation
An RNN arbitration model is trained on user data collected under shared control to blend human and robot commands in teleoperated pick-and-place tasks, with preliminary virtual-environment results compared to handcrafted baselines.