Abstract
Emotions, fundamental to human experiences, pose a persistent challenge in quantification due to their subjective and complex nature. Existing methods usually limit emotional quantification to two dimensions, failing to comprehensively capture the entirety of emotional fluctuations. In this research, a model termed as the "three-dimensional emotion space" (TDES) is introduced to elucidate and quantify emotions via pupil diameter (PD) and grip strength (GS) analysis, while introducing innovative metrics to discern emotional states. Specifically, an emotion induction experiment utilizing auditory and visual stimuli was first conducted to evoke three distinct emotional states (startled, stressed, amused), while collecting pupil diameter and grip strength samples. Kalman filtering and Wavelet denoising were employed to preprocess the raw data. Temporal and frequency domain features were extracted and mapped onto a three-dimensional coordinate system, resulting in a sequence of discrete points that outlined the trajectory curves and classification space of emotions. The results suggest the potential and feasibility of a novel method for quantifying emotions in a three-dimensional manner through the measurement of physiological signals, offering new perspectives in the field of emotion research.
Funding Information
  • National Key Research and Development Program of China (Grant No.2020YFB1313600)

This publication has 49 references indexed in Scilit: