<ul><li>Recent works in Task and Motion Planning (TAMP) show that training control policies on language-supervised robot trajectories with quality labeled data improves task success rates.</li><li>A framework is presented to decompose trajectory data into temporally bounded and natural language-based sub-tasks.</li><li>An algorithm named SIMILARITY is introduced to measure the temporal alignment and semantic fidelity of language descriptions in sub-task decompositions.</li><li>The framework demonstrates high scores for both temporal similarity and semantic similarity, above 90%, compared to a randomized baseline of 30% in multiple robotic environments.</li></ul>

Temporal and Semantic Evaluation Metrics for Foundation Models in Post-Hoc Analysis of Robotic Sub-tasks

Discover more