A joint neural network framework learns state predictions, action predictions, and lifted action models from unsupervised image sequences, using MILP optimization to enforce logical consistency and avoid prediction collapse.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning Lifted Action Models from Unsupervised Visual Traces
A joint neural network framework learns state predictions, action predictions, and lifted action models from unsupervised image sequences, using MILP optimization to enforce logical consistency and avoid prediction collapse.