Driver’s Speech Emotion Recognition for Smart Cockpit Based on a Self-Attention Deep Learning Framework

In-Vehicle Sensing for Smart Cars

“…Recent work [147] presented a review of the existing research on in-vehicle emotion sensing, which, according to the information adopted to sense emotion, can be divided into biological signal based (e.g., ECG, Heart rate and blood pressure, etc.) [235]- [242], speech signal based [243], [244], facial expression based [245]- [251], behavior based [252], and those using the combination of different features [148]. Biological signals-based methods can achieve good accuracy because those signals are directly related to the physiological response of a human being under different emotions.…”