Driver behavior analysis through speech emotion understanding

Driver behavior is indeed one of the major factors contributing to high number of motor vehicle accidents. Due to the fact that human behavior is always influenced by emotion and emotion can be detected through speech, we attempt to find correlation between driver behavior state and speech emotion to analyze driver behavior. This understanding is important to facilitate the development of driver emotional indicator system that can act as some kind of warning system to prevent accidents. Experimental results show potential for driver behavior state detection particularly for sleepy state based on speech emotion recognition approach coupled with fundamental understanding of affection space model. These findings surged us to propose an alternative approach of speech emotion profiling that complement the research mainstream of driver behavior and speech emotion recognition.