A smart hoodie integrally knitted by multi-functional electronic fibers for human emotion recognition and regulation

Abstract

Emotions play a vital role in human physiology and behavior, while emotional disturbances can significantly impact physiological and psychological health. Early detection and effective regulation of emotional states are thus critical for mental health management. However, conventional assessment methods primarily rely on subjective self-report questionnaires and behavioral evaluations, which limit their accessibility and reliability for non-expert users. Recently, wearable technologies, owing to their low cost, portability, and non-invasiveness, offer a promising platform for emotion recognition. Here, we present a smart hoodie that integrates multi-functional electronic fibers with machine learning algorithms to objectively detect negative emotional states and provide feedback. Given the correlations between emotional states and physiological signals, such as electrochemical markers in biofluids and electrophysiological activity, the integrally knitted hoodie can continuously acquire multimodal physiological signals, including electrocardiogram (ECG), electrodermal activity (EDA), and respiration (RESP). Leveraging these signals, the system achieves negative emotion classification accuracies of 77% for anger and 68% for sadness. Furthermore, by integration of luminescent fibers enables real-time optical feedback, forming an integrated system prototype for emotional self-regulation. This study provides an intuitive and interactive wearable platform for sensing and regulating negative emotions.