This repository deals with the issue of constructing a pddl action model out of AI agents videos, such that we can determine which fluent is holding in the current state via using image and object classifiers on the videos's frames, to feed the resulting action model with legality of performing actions and the result of applying them, so we can imporve the action model in case we have a partial one. (( I need to write down this explanation better ))
shaked-sapir/VIP-vision-PDDL
This repository deals with the issue of constructing a pddl action model out of AI agents videos,
Python