Towards Multimodal Driver’s Stress Detection - Digital Signal Processing for In-Vehicle Systems and Safety - page 4

Second, an analysis of speech production parameters in three cognitive load
scenarios is conducted, and a speech-based stress classifier is introduced. Third, a
classifier operating on CAN-Bus signals is proposed and evaluated.
1.2 UTDrive Corpus, Data Subsets,
and Transcription Protocols
The data collection vehicle is a Toyota RAV4 equipped with the following sensors
(illustrated in Fig. 1.1 ):
￿ Two CCD cameras for monitoring the driver and the road scene through front
windshield
￿ Microphone array (five mics) to record driver's speech as well as noise
conditions in the vehicle
￿ A close-talk microphone to obtain driver's speech with reduced noise content
￿ Optical distance sensor to obtain headway distance between equipped vehicle
and other vehicles in traffic
￿ GPS for location tracking
￿ CAN-Bus OBD II port for collecting vehicle dynamics: vehicle speed, steering
wheel angle, gas and brake inputs from driver
￿ Gas/brake pedal pressure sensors to collect information concerning pressure
patterns in car-following and braking behavior
Fig. 1.1 Instrumented data collection vehicle: UTDrive