Dual‐Modality Flexible Calorimetric Sensor for Speech Emotion and Human Motion Recognition

Abstract Human motion and speech constitute fundamental dimensions of human activities. Conventionally, these modalities are monitored separately using accelerometers and acoustic sensors, thereby introducing complexities in structural design and system integration. Here, a dual‐modality flexible calorimetric (DMFC) sensor is introduced, which is capable of monitoring speech through acoustic particle velocity and human motion‐induced airflow variation. The DMFC sensor is capable of effectively recording frequency information below 1000 Hz in human speech signals while demonstrating high sensitivity, strong linearity, and good reliability in measuring motion‐induced airflow. It enables precise detection of respiratory events, speech emotions, and body movements. When coupled with deep‐learning models, the sensor achieves 99.3% accuracy in distinguishing cough, breath, and speech emotions, and 97.0% accuracy in classifying five representative human motions. This compact, ultrathin, lightweight sensor offers significant potential for advancing healthcare and disease prevention in the era of population aging.