Learning Lifted Action Models from Unsupervised Visual Traces
Abstract
Efficient construction of models capturing the preconditions and effects of actions is essential for applying AI planning in real-world domains. Extensive prior work has explored learning such models from high-level descriptions of state and/or action sequences. In this paper, we tackle a more challenging setting: learning lifted action models from sequences of state images, without action observation. We propose a deep learning framework that jointly learns state prediction, action prediction, and a lifted action model. We also introduce a mixed-integer linear program (MILP) to prevent prediction collapse and self-reinforcing errors among predictions. The MILP takes the predicted states, actions, and action model over a subset of traces and solves for logically consistent states, actions, and action model that are as close as possible to the original predictions. Pseudo-labels extracted from the MILP solution are then used to guide further training. Experiments across multiple domains show that integrating MILP-based correction helps the model escape local optima and converge toward globally consistent solutions.
Cite
@article{arxiv.2604.19043,
title = {Learning Lifted Action Models from Unsupervised Visual Traces},
author = {Kai Xi and Stephen Gould and Sylvie Thiébaux},
journal= {arXiv preprint arXiv:2604.19043},
year = {2026}
}
Comments
Accepted to the 36th International Conference on Automated Planning and Scheduling (ICAPS-26)