Multimodal Features for Detection of Driver Stress and Fatigue: Review

Driver fatigue and stress significantly contribute to higher number of car accidents worldwide. Although, different detection approaches have been already commercialized used by producers (and third party companies), research activities in this field are still needed order increase the reliability these alert systems. Also, context automated driving, driver mental state assessment will be an important part cars future. This paper presents state-of-the-art review for evaluation. We describe details various signals (biological, video) derived features tasks we discuss their relevance advantages. In make complete, also datasets, acquisition systems experiment scenarios.