Objective This study aims to explore speech as an alternative modality for human activity recognition (HAR) in medical settings. While current HAR technologies rely on video and sensory modalities, they are often unsuitable for the medical environment due to interference from medical personnel, p…