Qualitative Study on Dense-Labeling Model Generalization for Semantic Concept-Augmented Learning and Post-Hoc Analysis for Time Series Classification

A qualitative study of dense-labeling model generalization for time-series concept identification on the UCR archive for the paper "Semantic Concept-Augmented Learning and Post-Hoc Analysis for Time Series Classification". This work will be presented at XAI-2026, to be held in Fortaleza, Brazil.