TransNet extends QMDP-Net with a state-classification module to learn and plan under non-uniform dynamics in POMDPs, showing better policy quality and learning efficiency on robot navigation benchmarks.
Value iteration networks.Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence, Aug 2017
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
-
Partially Observable Planning and Learning for Systems with Non-Uniform Dynamics
TransNet extends QMDP-Net with a state-classification module to learn and plan under non-uniform dynamics in POMDPs, showing better policy quality and learning efficiency on robot navigation benchmarks.