In-vehicle affect detection system: Identification of emotional arousal by monitoring the driver and driving style
Department of Cognitive and Learning Sciences
There have been clear needs to address the impact of driver emotions such as anger and happiness on aggressive or distracted driving behaviors. To tackle this issue, we have developed an affect detection system for identifying a driver's emotional arousal, including the driver's physiological data and vehicle's kinematic data. Multimodal sensors are wirelessly connected to a smartphone and then, all the driver and driving data are displayed on our Android application in real-time. With the benefits of this multimodal, portable, non-intrusive, and cost-efficient system, subsequent experiments were designed to test and improve the system. After identifying significant features, various machine learning algorithms will be used to model a driver's emotional states. Our final goal is to develop an optimized classifier of specific emotional states including arousal and valence. We hope that we can spark lively discussions on driver emotions at AutoUI and use the feedback to improve our system.
Adjunct Proceedings - 10th International ACM Conference on Automotive User Interfaces and Interactive Vehicular Applications, AutomotiveUI 2018
In-vehicle affect detection system: Identification of emotional arousal by monitoring the driver and driving style.
Adjunct Proceedings - 10th International ACM Conference on Automotive User Interfaces and Interactive Vehicular Applications, AutomotiveUI 2018, 243-247.
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/15210