Pre-Trained Foundation Model representations to uncover Breathing patterns in Speech