Emotional Adaptive Vehicle User Interfaces, Moderating negative effects of failed technology interactions while driving
Published in ‘Adjunct Proceedings of the 4th International Conference on Automotive User Interfaces and Interactive Vehicular Applications (AutomotiveUI 2012)’, 2012
Abstract
Automotive Natural User Interfaces have the potential to increase user experience providing intuitive interactions for drivers. However, in the complex setting of a driving vehicle, failed interactions with in-vehicle technology can lead to frustration and put drivers in a dangerous situation. This paper evaluates the possibility of applying emotion recognition to vehicular spoken dialogue systems in order to adapt the dialog strategies, in error recovery scenarios. An emotional taxonomy is developed for the interactions with a conversational vehicular application, the Voice User Help. The positive results of the performance of VUH emotion recognizer support the creation of real-time classification of the user emotional state, which serves as basis to emotional reappraisal dialog strategies that mitigate negative effects on the driver’s cognitive load and driver performance.
Key Contributions
- Developed the Voice User Help system
- Created a VUH-specific emotional taxonomy
- Built a real-time emotion recognizer
- Proposed adaptive emotional dialog strategies
- Used emotion for error recovery